You are on page 1of 316

Form Approved

OMB No. 0704-0188

Report Documentation Page

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it
does not display a currently valid OMB control number.



AUG 1999

00-08-1999 to 00-08-1999



Confrontation Analysis. How to Win Operations Other Than War







Office of the Assistant Secretary of Defense (OASD),Command &

Control Research Program (CCRP),Washington,DC,20301





Approved for public release; distribution unlimited


The original document contains color images.










19a. NAME OF


Standard Form 298 (Rev. 8-98)

Prescribed by ANSI Std Z39-18

About the CCRP

The C4ISR Cooperative Research Program (CCRP) has

the mission of improving DoDs understanding of the
national security implications of the Information Age.
Focusing upon improving both the state of the art and
the state of the practice of command and control, the
CCRP helps DoD take full advantage of the opportunities
afforded by emerging technologies. The CCRP pursues
a broad program of research and analysis in information
superiority, information operations, command and
control theory, and associated operational concepts that
enable us to leverage shared awareness to improve the
effectiveness and efficiency of assigned missions. An
important aspect of the CCRP program is its ability to
serve as a bridge between the operational, technical,
analytical, and educational communities. The CCRP
provides leadership for the command and control
research community by:

articulating critical research issues;

working to strengthen command and control
research infrastructure;
n sponsoring a series of workshops and symposia;
n serving as a clearing house for command and
control related research funding; and
n disseminating outreach initiatives that include the
CCRP Publication Series.

This is a continuation in the series of publications

produced by the Center for Advanced Concepts and
Technology (ACT), which was created as a skunk
works with funding provided by the CCRP under the
auspices of the Assistant Secretary of Defense (C3I).
This program has demonstrated the importance of
having a research program focused on the national
security implications of the Information Age. It develops
the theoretical foundations to provide DoD with
information superiority and highlights the importance
of active outreach and dissemination initiatives
designed to acquaint senior military personnel and
civilians with these emerging issues. The CCRP
Publication Series is a key element of this effort.

Check our website for the latest CCRP activities and publications.

DoD C4ISR Cooperative Research Program

Senior Civilian Official and Chief Information Officer, OASD (C3I)
Mr. Arthur L. Money
Director, Performance Assessment
Dr. Margaret E. Myers
Director, Research and Executive Agent for CCRP
Dr. David S. Alberts

Opinions, conclusions, and recommendations expressed or implied

within are solely those of the authors. They do not necessarily
represent the views of the Department of Defense, or any other U.S.
Government agency. Cleared for public release; distribution unlimited.

Portions of this publication may be quoted or reprinted without

further permission, with credit to the DoD C4ISR Cooperative
Research Program, Washington, D.C. Courtesy copies of reviews
would be appreciated.
Library of Congress Cataloging-in-Publication Data
Howard, Nigel.
Confrontation analysis : how to win operations other than war /
Nigel Howard.
p. cm. - (CCRP publication series)
Includes bibliographical references.
ISBN 1-893723-00-3 pbk
1. United StatesArmed ForcesOperations other than war.
I. Title. II. Series.
UA23.H568 1999
August 1999

How to Win
Operations Other Than War

Nigel Howard

Table of
List of Tables and Figures ...................................... iii
Acknowledgments ................................................. vii
Preface ................................................................... ix
Chapter 1 ............................................................... 1
Chapter 2 ............................................................. 21
Chapter 3 ............................................................. 61
Chapter 4 ............................................................. 95
Chapter 5 ........................................................... 133
Chapter 6 ........................................................... 155
Chapter 7 ........................................................... 189
Chapter 8 ........................................................... 215
Chapter 9 ........................................................... 253
Chapter 10 ......................................................... 275
Bibliography ........................................................ B-1

List of Tables
and Figures
Table 1. Confrontation between Allied Forces
and Rebels ............................................................. 9
Figure 1. Traditional relationship between
politicians and military .......................................... 14
Table 2. UN commander confronts ethnic militia .. 23
Table 3. Local commander confronts roadblock .. 25
Figure 2. The six phases of conflict resolution ..... 30
Table 4. The Drunk, His Wife, and His Friend ...... 74
Table 5. Status of Northern Ireland peace talks
when Tony Blair took over .................................. 119
Table 6. Status of Northern Ireland talks after
Blair shifted position ........................................... 124
Table 7. Northern Ireland options in
1993-1994, showing in detail the position taken
by the British and Irish governments in the
Downing Street Declaration................................ 141
Table 8. Moryan government no longer wants a
cease-fire ........................................................... 167
Table 9. The general has thought up a new
card, Blame the government ............................ 175


Table 10. After the Moryan president has

accepted the UNFORMOR position ................... 180
Table 11. Bosnian government refuses to
discuss a cease-fire............................................ 221
Table 12. Dilemmas facing characters in
table 11 .............................................................. 226
Table 13. What made UNPROFOR angry?
The situation preceding table 11 ........................ 228
Table 14. The dilemmas facing characters in
table 13 .............................................................. 229
Table 15. Ultimatum to the Bosnian Serbs ......... 233
Table 16. Grand strategic pressure on the
Bosnian Serbs .................................................... 237
Table 17. Ultimatum to the Bosnian Serbs with
context cards added from grand strategic
model ................................................................. 241
Table 18. Dilemmas in table 16 .......................... 243
Table 19. Dilemmas in tables 15 and 17 ............ 244
Table 20. Real pressure brought to bear at last
on the Bosnian Serbs ......................................... 247
Table 21. Dilemmas in table 20 .......................... 248
Table 22. Resolution (deceptive) of the
operational drama, with grand-strategic context 250
Table 23. Ultimatum to the Bosnian Serbs
with threatening grand strategic context............. 264
Table 24. Dilemmas facing characters in
table 22 .............................................................. 269


Figure 3. Computer screen showing briefing

on character role for Bosnian government ......... 283
Table 25. Confrontation between representatives
and suppliers ...................................................... 297


his book is based on research commissioned by

the UK Defence Evaluation and Research Agency
(DERA) at the beginning of 1997 into how
Confrontation Analysis might be applied to Peace
Support Operations. The idea of turning the research
into a book came from Dr. David S. Alberts, Director
of Research and Executive Agent for the C4ISR
Cooperative Research Program. For this I am most
grateful to him.
Andrew Tait (now with SAIC, then with DERA) was
responsible for suggesting the original research
project. His suggestion was taken up by George
Brander and Paul Willis of DERA, while Peter MurrayJones also took a keen interest. Many persons with
knowledge or experience of peace support managed
to find time to assist with the research, including
MajGen Pennefeather (Commandant-General of the
Royal Marines), Brig Alastair Duncan, Col Robert
Stewart (Retd), LtCol Philip Wilkinson, Brig B.C.
Lambe, Prof K.C. Bowen, Michael Codner and Edward
Foster (Royal United Services Institute for Defence
Studies), Prof James Gow, Indjana Harper and Vesna
Domani-Hardy. Capt David Fifield helped to make
military contacts, and Brig R. Lambe (Retd), Steve
Lea, George Rose and Graham Mathieson were
among many at DERA who helped by responding to
the ideas. The basic ideas have themselves been
worked out in collaboration with Peter Bennett (now
at the UK Department of Health), Prof Jim Bryant and


Prof Morris Bradley, as well as through interactions

with numerous academic and business colleagues
over many years.
Finally, I must express my gratitude to Dr. Richard E.
Hayes and Richard Layton, Evidence Based Research,
Inc., and their staff, Lynne Jennrich and Margita Rushing
for their work on drafts of the book, and Meg Rittler for
her work on the figures and cover design.

British Crown copyright 1998/DERA. Published with

the permission of the Defence Evaluation and
Research Agency on behalf of the Controller of HMSO.



his book presents a simple idea. A Peace

Operations campaign (or Operation Other Than
War) should be seen as a linked sequence of
confrontationsin contrast to a traditional, warfighting
campaign, which is a linked sequence of battles. The
objective in each confrontation is to bring about certain
compliant behavior on the part of other parties, until
in the end the campaign objective is reached. This is
a state of sufficient compliance to enable the military
to leave the theater.
If this simple idea is accepted, we can show how the
new technique of Confrontation Analysis (derived from
Game Theory via a development called Drama Theory)
can be applied. Thus we can show how to win an
Operation Other Than War.
Since this book was written, further research carried
out in the Bosnia theater has clearly revealed that
SFOR commandersfrom platoon commanders to
the overall theater commanderare doing it already.
They are winning confrontations, or campaigns, made
up of linked sequences of confrontations on a dayto-day basis.
They are doing it, however, without a clear, uniform
system of concepts specifically designed for a
confrontational campaign. Using practical good sense,
they are instead taking doctrinal concepts developed
primarily for warfighting, and adapting them for use in
confrontations. For example, they are using concepts


of artillery targeting to plan how to target

noncompliant parties (i.e., a local Mayor and police
chief who are refusing to provide security for returning
refugees from a different ethnic group).
Such common-sensical adaptation of standard
warfighting systems and concepts is admirable. And
it works.
We believe, however, that a system that treats
confrontations as confrontations, distinguishing them
from battles both conceptually and in terms of
planning procedures, will enable striking
improvements to be made so that our forces become
still more effective. In particular, it will make it possible
to use the powerful techniques of Confrontation
Analysis described in this book.
These are ideas that still have to win acceptance. This
book aims to lay them out for your consideration.

Chapter 1
The Need to
Operations Other
Than War
War is a mere continuation of politics by other

rom 1945 until 1989, U.S. and Allied defense

forces took as their first priority preparations for
a superpower conflict that never actually occurred.
Thus billions of dollars were spent in pursuit of a
theoretical construct.
Since 1989, the political assumptions on which this
construct was based have shifted. High-intensity
superpower conflict, the first and most important
contingency we had to prepare for, vanished from the
immediate agenda because only one superpower
remained. The result was that instead of a conflict
conceived and planned for using theory alone, defense
forces now must prioritize lesser threats of which they
have real experience.
Although this is real experience, as distinct from a
theoretical deduction, it is hard to make sense of in
traditional military terms. To the theorist using

Confrontation Analysis

traditional models, it is messy and disappointing. This

has put the defense community in a difficult position;
it has to ask for and use public money in pursuit of a
vision, persuasive for peace, decisive in war,
preeminent in any form of conflict (Joint Chiefs of Staff,
1997), whose conceptual and doctrinal underpinnings
it needs to clarify.
What is this messy, real-world experience?
Traditional Clausewitzian missions do come up, as in
the case of the Gulf War (Operation Desert Storm).
But the traditional character of even this mission
seems, in retrospect, less than obvious. Traditional
military objectives were, in the event, attained easily,
with minimal loss of life to the Allies and frightful losses
to Iraq. Yet the political problem was poorly resolved.
Other missions are not at all Clausewitzian. Peace
operations in general (see Alberts and Hayes [1995],
Maxwell [1997], and Wentz [1997]) have to do with
political stability and humanitarian assistance, rather
than physically compelling an enemy to submit to our
will. Objectives in these missions change and develop
as the mission goes on. Cooperation is required
between military forces and non-governmental
organizations that share responsibility for mission
objectives and must take them over when the military
leaves. Economic, political, and psychological
pressures are important. Military commanders must
negotiate with conflicting parties rather than fight them.
Despite these non-Clausewitzian characteristics, one
point made by Clausewitz (1968, 1st edition 1832) is
more relevant than ever. War and war preparations

Chapter 1

are political acts, whose nature varies with their political

background and aim.
Thus, to understand war in the era beginning after the
Cold War, and to begin to provide better conceptual
underpinning for defense budgets and plans, we will
try to understand the new political structures that
characterize this era.

The New World Order

What are defense forces now defending?
To be frank, what happened after 1989 was the
economic and military triumph of a single, unified world
system that U.S. President George Bush called the
New World Order (Bush, 1990). This system makes
surprisingly specific and detailed demands:

that all must be capitalist, democratic,

tolerant, non-racist, and non-sectarian; allow
equal opportunities and freedom of speech and
the press; protect human rights; not let
governments greatly incommode world trade
or capital movements; and be peaceful except
when enforcing these demands.
We admit, of course, that no sooner had Bush uttered
the words New World Order than the concept was
denounced as an absurd chimera by every
commentator. What is significant is that all recognized
at once what he meant and were able, unanimously
and without coordination, to agree as to what it was
they considered absurd. Their outraged denials,
repeated at intervals ever since, merely gave greater
substance and wider dissemination to the concept,

Confrontation Analysis

helping to make it the powerful coordinating framework

it has become.
The coordinating power of this universal understanding
is immense. When the United States and its allies seek
to impose peace and order anywhere in the world, all
know more or less what kind of order they are
demanding. Moreover, there is a definite tendency for
populations everywhere to demand it for themselves,
despite local efforts to persuade them not to.
Russia or China may still provide centers of resistance
to the New World Order. Their ability to do so is
doubtful, and decreases as they become more
dependent upon and involved in the system. As they
and others join it, the system will of course evolve. It
is essentially dynamic; however, fundamental change
can be expected to be continuous and evolutionary in
the manner of capitalist development, rather than
violent and disruptive.
The system now called the New World Order can be
seen, in retrospect, to have been growing within the shell
of the old, bipolar world since 1945. The old Clausewitzian
assumptions that war takes place between more-or-less
equal nations is carried out by military professionals
leading citizen soldiers, and is largely military in character
as distinct from political, held true exceptionally, if at all,
from 1945 until 1989. This is because, under the guise
of the Free World combating Communism, Bushs New
World Order was developing.
Now that the system has emerged into the open, how
does it increasingly enforce its disciplines?
Minor, particularly economic transgressions are
punished by economic sanctions. Transgressions that

Chapter 1

are not only violent but also sufficiently gross, wellpublicized, or economically damaging, such as in the
Gulf, Bosnia, Ireland, or Oklahoma, evoke forceful
responses. Defense forces are used then.
Among forceful interventions, the Gulf War generally
was counted a success, Somalia a failure, and Bosnia
a mixture of failure followed by the hope of eventual
success after the United States came in and got tough.
Northern Ireland has taken a turn for the better. The
threat of internal terrorism (e.g., the Oklahoma bomb,
the Tokyo subway gas attack) resembles an ongoing
war against an ever-varying enemy, with new methods
of attack continually met by new methods of defense.
The Arab-Israeli problem seems incapable of
resolution as long as the United States, because of
internal political divisions, remains unable to intervene
with sufficient force.

How Can the Strong Defeat the Weak?

Even over the Gulf War, counted a success, questions
are asked. At the end, why did Allied forces refrain
from unseating Saddam Hussein? Rogers (1997)
suggests that it was because the Allies feared
Saddams biological weapons, which he had delegated
to the control of his commanders in the event that
overall central command was lost. Although we do
not believe this, it shows the kind of power that might
potentially be wielded by a player that has been
comprehensively defeated.
Somalia and Bosnia expose in other ways the
inadequacy of conventional military responses to new
kinds of threat. Internal terrorist threats show this more
clearly. Consider not only the Oklahoma and World

Confrontation Analysis

Trade Center bombs, the IRAs economic targeting, and

the Tokyo subway attack, but the highly successful
destruction of the Bank of Sri Lanka by Tamil Tigers. In
each case, modern technology and terrorist or guerrilla
tactics have enabled the weak to take on the strong.
The problem of defense in the modern world is the
paradoxical one of finding ways for the strong to defeat
the weak. Obviously, this should not be an impossible
task; however, doing it with maximum effectiveness
and minimal loss of life and treasure does require a
new approach.

Confrontations, Not Battles

What kind of success can the weak aim for against
the strong?
The weak can never hope to defeat the New World
Order by fighting battles, the predominant mode of
warfare that defense forces prepared for for 40 years.
At most, the weak can hope for single, isolated victories
against local outposts. Some may see these as
significant because they believe the system is so
unpopular that their attacks will be copied and so fragile
it will then collapse. Others aim, more realistically, to
get away with something the system forbids: to commit
genocide against local ethnic neighbors (Rwanda,
Bosnia); to annex or transform a nation without a
democratic mandate to do so (Iraq, the IRA before
Easter 1998); or to grab weapons to loot and rape
neighbors (urban riots, Albania).
In opposing such rebels, the systems guardians
must make or reject demands. They must say either:
You must give up (genocide, aggression, etc.)

Chapter 1

or We refuse to (publish your manifesto, release

your comrades from prison, etc.). In either case,
defense forces must sustain and win not a battle
but what we may call a confrontation. This is a
situation in which victory consists of compelling,
persuading, or inducing others to submit to our will
without, if possible, using violence, although
violence, as well as other threats and inducements,
must be a credible part of our armory.
Forty years of preparation for the worlds greatest battle
have not equipped defense forces for winning
confrontations. Our aim in this book is to begin to
remedy this comparative lack of preparation.
We propose to investigate confrontations in general
using the technique of confrontation analysis, by
which the commander of a peace operation may plan
and execute a strategy for fighting and winning it.
We would like to claim, in fact, that by using this
technique it is possible to take a more logical, ordered,
and defensible approach to this problem than to the
analogous problem of winning battles.
This is not because emotion, friction, irrationality, and
the Clausewitzian fog of war are less present in
confrontations than they are in battles. On the contrary,
they are found in all types of conflict. What we claim is
that we have a method for analyzing these and other
factors systematically and scientifically.

Confrontation Analysis

A General Model
Confrontation analysis does not apply a general model
indiscriminately to every confrontation. On the contrary,
we model each confrontation separately, picking out
and taking advantage of its special features.
However, to show what a confrontation is, table 1 sets
out a simple, general model of one taking place
between the Allies and certain unidentified Rebels
against the New World Order. Here, the Rebel position
is that they should not be required to give in, but Allied
forces should concede their demands, the precise
nature of which, in this particular case, we do not
specify. The Allied position is that the Rebels should
give in without any concessions being made.
These positions are displayed in table 1 using the
metaphor of a card-table. This is simple. It works
as follows:
Each player (participating party) holds certain
cards, representing its yes/no policy options. In
table 1 players are listed on the left, with their
cards listed below their names (e.g., the
Rebels hold the cards Give in and Retaliate;
Allied forces hold the cards Concede and
Crush Rebels).
In general, a player can play any combination of
its cards (although some combinations may be
infeasible). In this manner a player chooses and
implements a policy.
After each player chooses a combination of its cards
to play, a column of cards laid out on the card-table
represents a projection of the situation as it would be

Chapter 1

Table 1. Confrontation between Allied Forces and Rebels.

determined by those policy choices. The choice by a

player to play a particular card is shown in table 1 by
a heavy, framed cell representing a card; the choice
not to play it is shown by a white cell. Column R, for
example, represents the future in which the Allies
concede the Rebels demands and do not crush them.
The Rebels, accordingly, do not give in (their demands
having been met) and do not retaliate either (action to
crush them not having been taken).


Confrontation Analysis

How Positions and Fallback Positions Are

Shown in a Card-Table
This, of course, is precisely the Rebels position (i.e.,
the solution the Rebels demand). The Rebels position
is shown by an appropriately labeled column.
The Allies position (i.e., the Allied solution, that the
Rebels give in and not retaliate, while the Allies neither
concede nor crush them) is shown in column A.
Column t shows the threatened future: What the
parties implicitly or explicitly threaten to do if their
positions are not accepted. The Allies say, If you dont
give in, well crush you. The Rebels say, If you try,
well retaliate.
This concept addresses an item of utmost importance
in a confrontation: what, in the last resort, a party
conveys to all and sundry that it will do if its demands
on the situation are not met. We call it the partys
fallback position.
Note that the players ways of conveying their fallback
position, which is a particular kind of conditional
intention, may be direct or indirect. A player may
sometimes convey its fallback position by denying it.
Imagine, for example, what it would mean if the
Russian president declared he has no intention of
invading a certain country, but that it should
nevertheless discontinue its anti-Russian policies.
The implicit, although explicitly denied threat in such
a statement would not necessarily be credible. It would,
however, be clearly made. The important point is that
at a climactic moment of truth in a confrontation, each

Chapter 1


partys fallback position (whether or not it is credible)

is clearly conveyed to the others.
In a card-table model, parties fallback positions are
represented by certain selections from their own
cards. Putting these selections together, we obtain a
future, called the threatened future. This is column t
in table 1. Here the Allied forces are implicitly
threatening to crush the Rebels (often merely bringing
armed forces into a theater is to implicitly threaten to
use them); the Rebels are threatening retaliation in
that case.

What This Model Represents

Many examples roughly fit our generalized model. In a
(highly simplified) model of the Northern Ireland
confrontation toward the end of the IRAs period of
armed struggle, the Rebels could be the IRA. The Allied
forces would be or the governments of Britain and
Ireland. Giving in might then mean accepting the
Anglo-Irish offer of negotiation in return for
decommissioning arms. Retaliation might mean a
continuing terrorist campaign. Conceding might mean
handing over power to Sinn Fein and the IRA. Crushing
the Rebels might mean continuing to fight terrorism.
This is a high-level example involving governments
and national movements as players. But the same
pattern appears at other levels. The first British troops
in Bosnia, for example, took on the task of clearing
roadblocks set up by different factions. We could fit
our model to the kind of situations they faced. The
Allied forces of our model would be the local force led
by a British colonel. Rebels would be the local ethnic
militia manning a particular roadblock. Giving in would


Confrontation Analysis

mean dismantling the roadblock. Conceding would

mean accepting its continuance. Crushing the rebels
might have meant using tanks to eliminate the
roadblock. Retaliation might have meant opening fire
on British troops.

Do We Know How to Do
How should a commander deal with a confrontation?
Does he know how to win it? What does winning mean?
Although individual commanders have been
successful in individual cases, there is no general,
trained military competence in this area. Yet our
argument is that winning confrontations is the clue to
conducting and winning peace operations. Just as a
traditional military campaign may be conceptualized
as a linked sequence of battles, so a peace operation
may, we suggest, be seen as a linked sequence of
confrontations. Continuing the comparison: winning a
traditional campaign consists of achieving overall
mission objectives through a sequence of battles, even
though objectives may not be fully met in each
particular battle, and some battles may be avoided
rather than fought. Winning a peace operation may
be defined similarly. We merely have to replace the
word battle with the word confrontation. The aim in
a peace operation is to achieve overall objectives
through conducting, and as far as possible winning, a
sequence of confrontations.
Yet compare the differences in preparedness between
fighting a traditional campaign and a peace operation.
Current military doctrine and training give excellent

Chapter 1


guidance as to how to conduct individual battles and

plan a victorious campaign through a whole sequence
of such battles. Complete, detailed instructions are
not, of course, given. Training and doctrine provide
guidelines, adaptable by the commander to meet
changing circumstances. In this way a commander is
taught how to plan a strategy and how to implement it
by devolving responsibility for its various elements
horizontally and vertically throughout his command.
By comparison, with the guidelines available for many
other human activities (e.g., setting up and running a
government department) these guidelines must be
judged to be highly scientific and effective.
Where do we find comparable guidelines on how to
fight and win individual confrontations and whole
sequences of them? Where is the training not only in
how to devise and follow a strategy, but in how to
devolve it to responsible units?
Figure 1 suggests an explanation for this comparative
lack of trained understanding. It depicts the traditional
relationship, implicitly assumed by Clausewitz and
other theorists, between politicians and the military.
Traditionally it was assumed to be the job of politicians
to resolve conflicts peaceably, if they could and wished
to do so (see top part of figure 1). When conflict
resolution failed or it was decided to use force instead,
the politicians directed the military accordingly,
instructing it to forcibly achieve political objectives (see
lower part of figure 1). The job for the military was to
conduct armed conflict and report back to politicians
on their progress. This and only this was what the
military was trained to do; actual, armed conflict
defined military professional competence.


Confrontation Analysis

Figure 1. Traditional relationship between politicians and military.

The development of nuclear weapons following World

War II caused one kind of change in this model: the
chain of command was modified to ensure that
politicians rather than generals made the decision, if
any, to escalate to nuclear war. That decision was felt
to be, as Clemenceau reportedly said of war in general,
too serious a business to be left to generals. Thus,
politicians began to intervene in decisions previously
left to the military. Such intervention became more
frequent as improved communications made it more
feasible. It was extended to more types of conflict as
the feeling grew that any use of force by a superpower
or ally of a superpower might start a process of
escalation; theoretical mechanisms accounting for
such escalation were spelt out by Kahn (1965). In this
way, nuclear weapons gave rise to a doctrinal concept
of limited war.

Chapter 1


At the same time, another kind of change was taking

place. Members of nuclear alliances no longer could
be challenged by forces strong enough to do so,
because of fear of nuclear escalation. The only
challenge that could still be mounted against them was
a challenge by the weak, such as the challenge of low
intensity conflict faced by Napoleons armies in Spain.
As we have said, meeting a challenge mounted by
the weak is essentially a matter of conducting a
sequence of confrontations. Objectives cannot be
achieved by force alone: the weak are not strong
enough and the strong face an enemy that avoids
decisive encounters by vanishing into the environment.
Although such operations are conducted by the
military, they can succeed only by reaching a political
solution, and this solution cannot be reached by highlevel decision makers alone because it must be
grounded in the hopes and fears of the foot-soldiers
fighting on behalf of the weak.
Consequently, the move to this kind of conflict had an
opposite result to the move to limited war; it meant
that the military was required to take over many of the
conflict resolution activities traditionally assigned to
politicians. Low-intensity conflict, it turned out, was too
messy, detailed, local, and political to be left entirely
to politicians to resolve.
Again, low-intensity conflict usually poses no direct
threat to the integrity of the powerful nations involved,
but only an indirect threat through the undermining of
international order. Because of this, the objectives of
low-intensity conflict are often far from clear. Often
objectives are ill-defined resultants of compromise at
the political level between different groups or different


Confrontation Analysis

nations in a coalition. As a result, the political directions

handed to the military are vague and ambiguous.
While limited war and low-intensity conflicts comprised
most of the reality faced by defense forces in the postwar era, it was imperative for them, while the Cold
War lasted, to train and study for the high-intensity,
total superpower conflict that never actually occurred.
This was their most serious task, on which national
survival and the future of the world ultimately
depended. The fact that this war never happened may,
paradoxically, be a result of the fact that it was so
soundly prepared for. This cannot be accounted a
failure. It was a success.
The price of this success was a comparative lack of
doctrine and training for limited war and low-intensity
conflict. This is the lack we hope to begin, at least,
to fill.

How To Do It, In Outline

Confrontation analysis, as said, proceeds by building
specific models incorporating the specific details of
each confrontation. It can be used to build models at
each level, strategic, operational, and tactical. These
models can be linked to enable a commanders
strategy to be devolved into strategies for each
subordinate level and linked to the strategies of players
related horizontally to his command, such as coalition
partners or non-governmental organizations.
Models so constructed will be strategic, looking forward
to winning a whole operation through a sequence of
confrontations. They will be readily capable of
changing to meet changing contingencies, and such

Chapter 1


changes will be readily propagated through the system

of linked models.
By presenting such a system in a simple, clear way,
accessible to all professionals involved, we would hope
eventually to provide doctrinal guidance and training
for the at-present, vaguely defined task of peace
support that is at least as good as that provided for
fighting battles.
We would propose that eventually a theater
commander in a peace support operation would have
analysts on his staff trained in confrontation analysis.
At each stage, starting with his first notification of the
mission he is tasked with, these specialists would
model the problem at each command level.
The commanders first need is, in general, to
understand the problem facing his political masters,
because understanding the intent of his commander
is necessary for him to understand his own mission.
To help him, analysts would model the peace mission
as part of the world-political problem and the specific
peace-support problem, as he sees it.
When a commander receives a specific mission,
perhaps in vague, nonspecific terms that result from
compromises between political actors, analysts would
work with him to model the specific confrontations he
is directly responsible for handling. Such analysis,
incorporating the commanders own assumptions,
should yield the two following results:
Establishment of a sequence of steps to apply
requisite pressure on other parties to bring them
into compliance with the commanders position
(i.e., mission objectives). We shall see how to


Confrontation Analysis

derive this sequence based on dilemmas (changeinducing points of stress) we and others face, and
how these are handled to achieve our objectives.
Invention of an immersive role-playing exercise
through which the commander and other officers
involved can digest and criticize the analysis,
rehearse their interactions with other parties, and
become familiar with other-party points of view.
Analysts would then work with lower-level
commanders and representatives of other
components to build devolved and horizontally linked
models. Both the analysis and the role-playing based
on it would be updated as the confrontation evolves.
The role-playing would be a kind of war-gaming, but
with a firm analytic basis.
This, in outline, is the system we hope eventually will
fill the present gap in doctrine and training for
Operations Other Than War (OOTW). Chapter 6 will
show how it might work. It contains the Frontline Play,
an attempt to dramatize the situation of an imaginary
commander of a peace support operation in a fictitious
country who decides to analyze his problem using
confrontation analysis.
Meanwhile, chapters 2 and 3 give a detailed account
of what confrontation analysis means.

Summary of Chapter 1
Following the end of the Cold War, defense priorities
generally have shifted to a new kind of mission, broadly
described as defending the New World Order against
Rebels who are militarily weaker. To understand this

Chapter 1


kind of mission, the concept of winning a campaign

by fighting a linked sequence of battles needs to be
replaced with that of winning an OOTW by conducting
a linked sequence of confrontations.
A simple, card-table model of a general confrontation
shows the position and fallback position of each party.
A commanders mission is to get all parties willing
compliance with his position. In doing this, he needs
to apply pressure on other parties in ways traditionally
thought of as being a politicians job. A professional
approach to this requirement is needed. It can be
supplied by confrontation analysis.

Chapter 2
Handling a Multilevel
The Six-Phase Model

inning an Operation Other Than War (OOTW),

we have said, is a matter of conducting and
winning a linked sequence of confrontations. The
question is: How does a commander form and
implement a strategy for this?
This is the key question. To begin to prepare an
answer, in this chapter we will look at the process a
confrontation goes through, leading it to being resolved
or breaking out into conflict. The commanders aim is,
of course, to have it resolved on his terms.
His strategy for achieving this is not primarily a matter
of physical activities, as it is with warfighting. It is a
matter of communication. At the same time, his
possession of a credible capacity for warfighting is
usually central to it.
As with warfighting, his strategy needs to be
implemented on many levels. He should be able to
give directives that link together the many
confrontations occurring on different levels and
implement them through a cohesive, multilevel
strategy. To make this point more clearly, we begin
with a concrete, albeit simplified, example.



Confrontation Analysis

Example: Removing Roadblocks

Table 2 shows a card-table model of the situation faced
by a UN commander who has been tasked to make a
certain ethnic militia force stick to an agreement it has
signed. This model differs from the prototype in chapter
1 (see table 1) in that both parties are taking the same
position. This does not, however, mean they have no
further problems. The numbers alongside players
names in this model represent their preference
rankings (the order in which they prefer the three
futures shown). The most preferred future for that
player is given the number 1, the next most preferred,
number 2, and so on.
We are making the following assumptions:
Assumption 1The agreement states that the
militia should cease ethnic cleansing (i.e.,
attacking villages and relocating members of
another ethnic community) and allow free
movement of people. The commanders cards
include the following options: (a) give or withhold
support and aid to the militia, which constitutes
the effective local government; (b) if the militia
disallows free movement, the commander may
force it by removing roadblocks and escorting
travelers; (c) if the militia flagrantly violates the
agreement, the commander may bombard the
militias headquarters, as he was doing until the
militia signed the agreement.
Because the militia has now signed the agreement, it
shares the commanders position that the agreement
be kept (column P); however, the commander
suspects that the militia does not actually intend to

Chapter 2


Table 2. UN commander confronts ethnic militia.

allow free movement, which would encourage

displaced villagers to return to their homes, explaining
the question mark placed on this card. Hence, the
confrontation is not really resolved. [Note: The word
intend needs clarifying here. In what sense does a
complex, multifarious organization such as an ethnic
militia intend anything? Answer: A players objectives
and beliefs are considered to be the result of internal
confrontations between internal factions (subplayers
belonging to and identifying with the player). In our


Confrontation Analysis

case, the militia leadership (one subplayer) may want

to allow free movement, but the leaderships followers,
who are local villagers operating locally (and constitute
another subplayer), do not. The commander, we
assume, suspects that the followers view may prevail.
In this sense, the commander suspects the militia
intends to defect from the agreement.]
Assumption 2The current threatened future
(column t) is not the threatened future whose
actual implementation (bombarding of the militia
headquarters) led to the signing of the
agreement. It is the future the commander is
now threatening to enter on (by implementing his
part of it) if the militia does not allow free
movement. Simultaneously the militia is
threatening to implement its part if the
commander does not give aid and support.
The term default future, attached to column d, is a
general term for the future now being implemented
through the parties current policies, and which will
therefore continue unless and until those policies are
changed. The militia has given up ethnic cleansing,
but it has not yet allowed free movement. The
commander has not started giving aid and support;
nor has he started forcibly removing roadblocks.
Note that the cards bombard militia HQ and resume
ethnic cleansing are included in the model although
they are not played in any characters position or
fallback position, nor in the default future. They are
there because they are active possibilities in players
minds. A card-table model should include all cards
the players consider relevant. It must include all cards

Chapter 2


needed to define players positions and the threatened

future; but it may include more, as in this case.
In summary, the UN commander has bombarded the
ethnic militia until it agreed to give up ethnic cleansing
and allow free movement. The commander then
promised to give the ethnic militia aid and support.
He has not done so yet, and the militia has not yet
allowed free movement; thus, the agreement has not
been fully implemented.

Linkages Between Different Levels

Next, to illustrate linkages between two levels of
command, table 3 models the situation of a local
battalion commander faced with a roadblock. The local
militias position (M) is that the roadblock stays. In light

Table 3. Local commander confronts roadblock.


Confrontation Analysis

of his commanders intent (as shown in table 2), the

local commanders position is that it be removed or
he will forcibly remove it (t). If the militia then fires on
his troops from surrounding hilltops, the commander
will call in air strikes (tI).
Note that the players are considering two possible
threatened futures. This is because the ethnic militia
has not made its fallback position clear (i.e., whether
it will fire on UN troops if roadblocks are forcibly
removed). This ambiguity indicates that the
confrontation is still in the process of building up to a
climax. At a proper climax, all positions are made clear.
The local commanders card-table is linked to his
commanders card-table; it represents the
implementation at the local level of the commanders
overall position and fallback position.
This statement is clearly true in a general way;
however, it is important that the exact manner of the
linkage be decided by the local commander based on
his particular circumstances. It is not determined in
any mechanical way by the commanders position.
What is true is that the commanders position should
imply that the local commander must make sure the
roadblock is removed and give a general idea of the
sanctions he can use: first, forcible removal, then
adequate punishment of any retaliation on the part of
the militia.
The linkage between these two levels is mutually
reinforcing. If the commander took up his position and
fallback position only at the level of his discussions
with the militia leadership, without making sure his
lower-level commanders implemented corresponding

Chapter 2


positions at their respective level, his position would

be far less credible. Similarly the credibility of the local
commanders position is enhanced by its being
supported by the position of his commander.

Conflict Resolution Is Not a Matter of Taking

Physical Action
Is not conflict resolution just a matter of the commander
issuing orders to all local commanders to forcibly
remove roadblocks and call in reinforcements if
necessary to counter opposition?
The answer is no. Dealing with a confrontation is not
just a matter of taking such-and-such actions,
contingent or otherwise, upon the actions of others,
although such actions may be a part of it. This is
shown, in our case, by the fact that forcible removal of
the roadblock is not part of the local commanders
position in table 3. His position is that the local militia
should remove it.
This is far from a trifling difference. Handling a
confrontation is a matter of resolving a conflict through
dialogue. The aim is that actions eventually taken
should be willingly agreed to by all parties, ideally seen
by each party as fulfilling its own objectives.
This would be ideal in the case we are examining. If
the local UN commander gets his way in this manner,
then the message going up from the grassroots to the
militia leadership will be, Agree to remove roadblocks!
Were in favor. Forcible removal of roadblocks, on
the other hand, would run the risk of increasing local
opposition to their removal, or cause grassroots
demands that the militia leadership resist this. It might


Confrontation Analysis

thus work against the UN operational commanders

mission objectives.
We are emphasizing, through this example, the
difference between physical action, including the use
of force, and resolving a conflict. The latter is not done
directly through force, but through dialogue conveying
the credible threat of force, leading, if successful, to
genuine agreement.
The difference must be stressed because it is often
obscured by the fact that dialogue itself may require
physical acts, even large-scale ones, to demonstrate
credibility. For example, warlike preparations, even
up to beginning to use force, may be necessary to
send the message that force will be used unless
certain conditions are agreed to. But this is still
dialogue, even though lives may be lost. It is dialogue
because the aim is to send a message rather than
directly to achieve objectives. This can be seen from
the simple fact (Alberts and Hayes, 1995) that the
element of surprise, always desirable when force is
used directly to achieve objectives, must be absent.
When forceful preparations, or force itself, are used
to send a message, it is essential for the other side
to know about them.
It appears then that force, although often an integral
part of conflict resolution as carried out by armies,
should be avoided if possible. To avoid using it, its
use must be made credible.
What exactly is going on here?

Chapter 2


The Six Phases of

Conflict Resolution
We need to explicate the place of force (or, more
generally, actually realized conflict) in the process of
conflict resolution. Figure 2 shows the general conflict
resolution process divided into six phases. We will
discuss these one by one.

Phase 1: Scene-setting
In this phase the problem to be resolved is set before
the parties by the context, by a higher authority, or by
a preceding confrontation. For a commander, this
might be when he is tasked with a mission, or when
new circumstances arise during a mission.
For the confrontation in table 3 we can be fairly specific.
Scene-setting consists of the roadblock set up by the
local militia and the receipt of information about it by
the local commander. These are the events that bring
them together in a confrontation.

The Informationally Closed Environment

As figure 2 indicates, scene-setting must logically set
up a so-called informationally closed environment
within which the confrontation may be resolved. This
concept is important. Issues between parties can be
resolved only on the basis of information available to
them at the time, and this information must remain
fairly stable during the course of negotiations.
Resolution is impossible if new, relevant information
continually upsets nascent understandings and
commitments. Following are two contrasting examples


Confrontation Analysis

Figure 2. The six phases of conflict resolution.

Chapter 2


that illustrate how demanding the requirement for

informational closure can be:
Recently a state came to a crisis because of
persistent feuding between the private armies of
two politicians. As a last resort, the army chief of
staff took over and locked the politicians in a
hotel room with phones cut off until they could
reach agreement. In doing so, he enforced
informational closure.
During Operation Market Garden (the Allied drop
on Arnhem in World War II), a reconnaissance
pilot observed German heavy tanks in an area
where British paratroopers were scheduled to
drop. This indicated the presence of an SS
panzer division; however, there was resistance
to this genuine new information. Senior Allied
staff had spent much time and effort agreeing to
battle plans on the basis of the best available
information. This naturally created a mind-set
according to which conflicting new information
should, if possible, be seen as incorrect. The
alternative, after all, might require lengthy
renegotiations of positions (in drama theory
terms) and redrawing of plans throughout large
organizations, which would have cost a great
deal in terms of time, resources, and trust
between players. The report was ignored.
The second example shows how informational closure,
if taken too far, can have pathological effects. It
nonetheless shows how necessary it is for conflict
resolution, in this case, resolution of internal
confrontations between Allied staffs. The point is that
negotiations require stability, at some level, in the facts


Confrontation Analysis

accepted as common ground between negotiating

parties. The operational need is to resolve
confrontations in a way that allows for flexible
responses to new information, while preserving mutual
confidence and understanding. Hopefully confrontation
analysis will help us to meet this need.

Phase 2: Buildup
In this phase, dialogue takes place between the
confronting parties to bring them into full confrontation.
As figure 2 indicates, in this manner parties take up
final positions within a common reference frame. This
needs explaining.

The Concept of Common Reference Frame

Although communication between confronting
parties does not require agreement between them,
it does require a degree of mutual understanding.
They must know the meaning of terms used by each
other and know that each other knows that they
know. The following analogies illustrate this: It is no
use for me to threaten you with a gun if you think it
is a popsicle. Nor can I bribe you with a popsicle if
you take it for a gun.
The game-theoretic term common knowledge is useful
here. It stands for what each party knows and knows
that others know and knows that others know that
others knowand so forth. Using this term, we can
say that confrontation requires common knowledge
of the assumptions underlying communications
between parties.

Chapter 2


Without such common knowledge, they cannot

adequately communicate threats and promises to each
other; this requires that each must not only understand
the other, but know that the other understands them,
and so forth.
This essential set of communications assumptions
shared between parties is their common reference
frame. It is this set of assumptions that tables 2 and
3, showing confrontations with a militia force, are
modeling. We consider a common reference frame
to be specified by designating a card-table, together
with an indication of how each player would rank
the various possible futures in order of preference.
These preference rankings can be conveyed by
assigning players priorities for the various futures,
as in our models.
The point here is that communication requires
common knowledge of what the other is talking about
(i.e., who the players are and what their cards are)
and of each players assumed preferences for
outcomes (e.g., without common knowledge of
preferences, what is meant as a promise may be
taken as a threat, or vice versa).
Following are key points concerning common
reference frames:
Simplicity is required for a common reference
frame. Only if a common reference frame is very
simple can each party be sure that it is common
knowledge (i.e., can really think, I know that you
know that I knowwhat I mean.). An example is
when U.S. negotiator Richard Holbrooke
confronted Bosnian Serb commander General


Confrontation Analysis

Mladic and Serbian President Slobodan

Milosevic in Dobranovci, near Belgrade, in
September 1995. Their common reference frame
was very simple. It contained two players, NATO
and Serbia. The cards under discussion were
NATOs continued bombing of the Bosnian Serb
army and the Serbian withdrawal of guns from
around Sarajevo. This, said Holbrooke, unless
that. All details that might distract from this
simple message receded into the background.
(See Silber and Little, 1996).
This could be considered an over-simplification
of a complex reality. The Bosnian government
and Bosnian Serbs were independent actors.
Milosevic recently had obtained a kind of
mandate to negotiate for the Serbs as a whole,
but on hearing Holbrookes demands, he said,
Come and tell Mladic yourself. Holbrooke
agreed; but Mladic was adamant. Holbrooke left,
threatening NATO action. Milosevic warned
Mladic, telling him that unless he agreed, NATO
would destroy his army. Mladic finally agreed.
A more complex model could be built; however,
any model can be made more complex by
including more detail. It also can be made
simpler by generalizing and omitting details.
Simple models are better representations of the
stripped-down common reference frames of
players at the climax of a confrontation than
ones that include distracting details. Such details
as those mentioned above would merely
illustrate our earlier point that a players
decisions are determined by confrontations

Chapter 2


between subplayers within that player (such as

the face-off between Milosevic and Mladic).
A common reference frame may or may not
represent parties actual views. This point too
is important. If one party succeeds in deceiving
another (e.g., by concealing the fact that it
possesses certain weapons) then the common
reference frame does not include the card, Use
those weapons, even though, if the truth were
known, both parties would include that card in
their considerations. Similarly, a party may
deceive another by pretending to have a weapon
it does not have, or pretending (falsely) to have
the will to use a particular weapon. Going
further, suppose that the second party is not
deceived by some such pretense, but pretends
to be. Then both parties will in fact share the
same frame, while communicating in terms of a
common reference frame believed in by neither
of them. If these parties resolve their conflict,
they do so within a common reference frame
known by at least one of them to be inaccurate.
The common reference frame, being
common, does not reflect the parties
differing values. Each party generally considers
itself to be in the right and thinks that others
values, and hence their way of describing the
situation, are wrong. Thus they often use
different terminology for elements of the
common reference frame. What one calls
freedom fighters another may call terrorists
and a third drug runners. A common reference
frame, in our sense of the term, still exists,


Confrontation Analysis

provided that all parties know what each other

means, and hence know what each other is
referring to using different terminology. Thus
they are referring to the same players and cards
(which for purposes of analysis we may name as
we like, using neutral terms that best suit the
party we are communicating with) and to the
same preferences, even though different parties
with different ideologies may give different
reasons for and evaluations of these.

Final Positions
As figure 2 indicates, something more is required for
the parties to build up (during the Buildup phase) to a
full confrontation. The parties must take up positions
and fallback positions within the common reference
frame, as shown in tables 2 and 3, and these too must
be common knowledge.
During the Buildup phase they may experiment with
various positions to see how the others react. The
Buildup phase comes to an end when the parties
adopt positions they consider final within what they
consider a final common reference frame. In the end,
if the parties are to resolve matters, they must take
a stand.
They then face what we call a moment of truth,
defined technically as a frame together with positions
and fallback positions for each party.
At this point, one of two things must be true. Either all
parties take the same position and, according to their
common reference frame, can trust each other to
implement it, or this is not the case.

Chapter 2


Suppose it is the case. Then the parties have solved

their problem as defined by their frame and their
objectives within it. A more detailed model, or a model
representing different issues, might show they have
differences. Within this frame, they have none. As
illustrated in figure 2, they therefore proceed next to
the Resolution phase.
This does not necessarily mean the parties are being
honest with each other. A solution according to their
proclaimed positions and common reference frame
may not be a true solution. One may be deceiving
another. Each may be deceiving all the rest. If there
is deception, it necessarily consists of each deceiver
acting as if it has reached honest, trusting agreement.
This is because successful deception requires the
deceiving party to behave, not in accordance with
the facts it believes to be true, but in accordance with
the pseudofacts it wishes to be accepted. The
behavior of such parties (to the extent that they are
successful) therefore will be indistinguishable from
the behavior of parties that are not deceiving each
other, but genuinely believe in the version of reality
they are projecting.
For example, the end period of the Nazi-Soviet Pact,
initially signed in August 1939, may be a case of mutual
deception. Although evidence of Soviet intentions is
lacking (see Davies, 1996, pp. 1000-1013), it seems
from their force dispositions that by June 1941 each
was preparing a surprise attack on the other. The Nazis
happened to strike first. Meanwhile, as far as their
discussions went, the pact stood. This case actually
takes the deception one twist further than we have
described, in that not only did they deceive each other,


Confrontation Analysis

they also disbelieved each other; however, each may

have mistakenly believed that it was believed.

Phase 3: Climax
We have seen that one possibility at the end of the
Buildup phase is that the parties find they have solved
their problem, in that the dialogue between them
records full, trusting agreement, even though,
because of deception, the agreement may not turn
out as expected.
What about the second case? Here parties either
disagree on the terms of a resolution or openly distrust
each others intention to carry it out. They then move
to the Climax phase as seen in figure 2. Here they
face what is correctly called a moment of truth;
something has to give. If it does not, they move into
the Conflict phase, which consists of each undertaking
to carry out its fallback position.
Fear of this outcome, which is in general not liked
by one or more parties, puts pressure on them to
change; therefore, something must give, and what
must give is either their fixed positions within the
given frame or the fixed frame itself. In other words,
to avoid the Conflict phase, they must change one
or more of the following:
ObjectivesTheir own or others objectives
(encapsulated in their positive positions)
ThreatsTheir own or others implicit threats of
what they will do if they do not get their objectives
(encapsulated in their fallback positions)

Chapter 2


BoundariesTheir own and others beliefs about

the boundaries of their negotiations (modeled by
the relevant set of players and cards)
MissionsTheir own or others general aims
pursued in the confrontation (resulting in the
preferences they attach to various combinations
of cards).
For example, look at the encounter described between
U.S. negotiator Holbrooke and the Serb leaders. At
the climactic moment of truth, the Serbs changed
position. They agreed to withdraw their weapons. Did
they also change their mission? Were they still
determined to cleanse Sarajevo of Muslims? Most
probably they were. The grudging nature of Mladics
change of position indicates that his concept of what
the Serb mission should be did not change; however,
Milosevic succeeded in overruling Mladic, and
Milosevic may by then have had a different concept
of the Serbs mission. In that case, we would call the
change a mission change as well as a change of
position, although the mission change would be seen
as precarious, depending on the balance of power
within the Serb coalition.

Emotion and Reason

How then is change brought about? The answer is by
means of emotion (positive and negative) and rational
debate (or arguments in the common interest).
Holbrooke must have expressed the anger of the
United States and the international community at the
Serb bombardment of Sarajevo. (Note that ranting and
arm waving are not the only ways of expressing anger.


Confrontation Analysis

An expressionless face in answer to anothers emotion

may suffice.) However expressed, such negative
emotion would have helped to make it credible that
he would bomb the Serb army into oblivion. Mladic,
too, must have expressed anger, signaling his
determination not to give in. At the same time, as a
skilled negotiator, Holbrooke may have expressed
positive emotion toward the Serbs contingent on
changing their mission. The negative emotion of fear
finally would have propelled the Serbian change.
Such emotions necessarily will have played an
essential role; however, emotion on its own is not as
effective as when it is supported by rational arguments
in the common interest, based on logic, and if possible,
including the production of evidence.
Holbrooke would have pointed out that neither side
wished to see the Serbs destroyed. (Note that it was
necessary to make this clear to the Serbians because,
had it been part of NATOs mission to destroy them,
the threat to do so would not have placed pressure on
them to withdraw, because NATOs mission would
have been to destroy them whatever they did.) For
his part, Mladic may have tried to appeal to a supposed
SerbianUnited States common interest in opposing
Muslim expansionism.
The function of rational arguments in the common
interest, using logic and evidence, is to build credibility.
Mere emotion can appear to be not serious because
its effect may be considered transient; however, if I
can convince you that my position and fallback
position, in addition to being invested with strong
positive and negative emotion, also are based on logic
and evidence and are in the common interest, they

Chapter 2


become very credible. This is because offering reasons

makes it harder for me to give in and easier for you.
Often, neither reason nor emotion, on their own, are
credible motivators; together, they are.
Rational common-interest arguments also tend, if
successful, to weld the confronting parties together
into what is seen as one party, pursuing a unified set
of interests. This is the converse of the fact, already
noted, that the preferences and attitudes of a player
are determined by confrontations between the
subplayers composing that player.
Subplayers, by continually solving their subconfrontations
within a player, in effect continually construct the player
to whom they belong as a purposeful entity able to take
decisions and, through its representatives, express
emotions. Similarly, when any confrontation is properly
resolved, good feelings between the parties and
sentiments of loyalty toward the agreement reached are
created. If organizational requirements also are met,
these feelings tend to create a single superplayer
comprised of the original players.
Some further points need to be made about emotion
and reason:
Warning and NotificationIf I do not need to
change my preferences to make my threat or
promise credible (because they are things I would
want to do anyway), my threat is more like a
warning and my promise more like a notification.
To convey this, I may show lack of emotion.
DeceptionConversely, to deceive you into
thinking that my threat is merely a warning, my
promise a mere notification, I may show lack of


Confrontation Analysis

emotion. For example, Holbrooke may have tried

to show the seriousness of the U.S. commitment
by deliberately showing no emotion.
IrrationalityIf I have no good reasons for
preferring to reward or punish you, I may make
my promise or threat credible by strong positive
or negative emotion accompanied by evidence
of irrationality so that you think, with delight or
horror, He is mad enough to do it! This is more
likely to be effective, the more immediate is the
possibility of implementing the threat or promise
because we know that emotions often are
temporary. For example, your anger is more
likely to be effective if your guns are pointing at
someone than if it will take 3 weeks to transport
them from the United States.

Phase 4: Conflict
If a change in positions or common reference frame
takes place in the Climax phase, it requires a return to
the Buildup phase, as shown in figure 2. This is
necessary to communicate the new configuration of
positions within the new reference frame and to
establish that they are common knowledge.
It is possible change still may not take place. Emotions
and the search for reason and evidence may not be
enough to overcome inescapable evidence (e.g., Hitler
has invaded Poland), unforesakeable values (e.g.,
Europe should be democratic rather than Nazi-ruled),
and unshiftable positions (e.g., Poland shall not be
occupied by Germany; if it is, Britain will declare war).
If change is impossible, by assumption the characters
must resort to their fallback positions; that is, move to

Chapter 2


the Conflict phase (e.g., World War II). Here the parties
must prepare for what we have called the threatened
future, the appropriate term for the conflict as imagined
or foreseen by the parties.
At this point, the parties split up (their bid to achieve
their position having failed), and each decides if it
now intends to implement the fallback position it has
implicitly or explicitly committed to. This phase is a
serious decision point. Internal subplayers belonging
to each player now may confront each other, some
demanding to carry out their commitment, others
unwilling to do so. Estimates on what other parties
will do are important to a party making this decision.
Will they carry out their commitments to the
threatened future? What alternative courses of action
will they take?
These internal confrontations also can be modeled
by the card-table method, using models linked to the
original overall table.
In Athens in May 1993, United States and British
envoys Cyrus Vance and David Owen, backed by
Bosnian President Slobodan Milosevic, persuaded
Bosnian Serb leaders to sign the VanceOwen plan
for dividing Bosnia while keeping it as one nation.
According to Vances right-hand man: Vance
suddenly tells them straight out that this is the lastchance caf, the U.S. Air Force is all prepared to turn
Bosnia and Serbia into a wasteland. This was the
threat Milosevic had been hammering home to them.
So the Bosnian Serbs signed, subject to the approval
of its parliament, which rejected the plan. For a time
after, U.S. officials tried to persuade Europeans to
agree to implementation of their threat. They failed.


Confrontation Analysis

This United StatesEuropean subconfrontation ended

with the United States shifting position and agreeing
not to anger its NATO allies by act[ing] alone in taking
actions in the former Yugoslavia. There was a
concomitant change in United States values (i.e., the
values it was pursuing in former Yugoslavia), initiated
by President Bill Clinton and his wife Hillary, who read
a book that persuaded them the Balkans were doomed
to violence. We interpret this as follows: In the Climax
phase of the West-versus-Serbs confrontation, the
Serbs decided to call the Wests bluff. The players
then moved to the Conflict phase, in which the Wests
decision (through the resolution of confrontation
between the United States and Europe) was to not do
as they had threatened to do (Silber and Little, 1996,
chapter 21).
Thus the Conflict phase may end in an Implementation
phase in which the threatened future is carried out or
one in which it is flunked. There is a third alternative.
There may be a return to the Climax phase if one or
more parties decides, as a result of internal
confrontations, that after all they are prepared to
change, and therefore request to reopen negotiations.
They may then cycle back from the Climax phase to a
renewed Buildup phase.
This happened in the example of the Holbrooke
MilosevicMladic confrontation of September 1995.
When Holbrooke left the meeting, negotiations moved
into the Conflict phase. Milosevic and Mladic then had
a subconfrontation as subplayers within the player
Serbs. This was resolved by Mladic caving in. The
Serbs then called Holbrooke back.

Chapter 2


The Use of Force

In this example, the Western threat was to use force. Is
a threatened future necessarily one of armed conflict?
At the political level, a threatened conflict may not
involve force at all. Economic sanctions, for example,
may be the only threat. The mere involvement of the
military tends to carry the implicit threat of the use of
arms, even though not necessarily their direct or
immediate use.
In classic United Nations peacekeeping operations, if
former combatants reject their position in favor of a
continuing cease-fire, the fallback position of the
United Nations is simply to withdraw UN troops. There
is no theater-level threat of United Nations use of force,
although this is threatened at lower levels if there are
attacks on UN troops, who are supposed to defend
themselves if attacked. Although the UN does not itself
threaten to use force at theater level, UN withdrawal
may lead to resumption of armed conflict between the
parties, as happened in May 1967, when the
withdrawal of UN troops was followed in June by the
Six-Day War (Alberts and Hayes, 1995, p. 18).

Phase 5: Resolution
If parties are to finally resolve their differences, rather
than fall into Conflict, according to figure 2, they must
cycle between phases of renewed Climaxes and
renewed Buildups until they converge on a common
position within a frame that ensures them of each
others trustworthiness. Each time they cycle through
the Climax phase, they risk moving into the Conflict


Confrontation Analysis

phase. If they successfully resolve their differences at

the Climax level, they move to the Resolution phase.
After the parties sense they have reached an
understanding sufficient to resolve their problem, they
must communicate to give mutual reassurance that
each understands what is required. This takes place
in the Resolution phase. Here details are discussed
and understandings translated into more or less
unambiguous commitments.
It is always possible that this consideration of details
and firming up of an overall understanding may reveal
flaws sufficient to differentiate the parties overall
positions or cause serious mistrust. That will
necessitate a return to the Buildup phase followed by
a renewed Climax phase, with the renewed danger of
conflict. To avoid this, the detailed discussions need
to be carried out in an atmosphere of mutual goodwill
(positive emotion) to encourage parties to see each
others viewpoint, find win-win solutions, and trust each
other; however, this general atmosphere cannot mean
that detailed differences are not explored. That is the
reason for this phase. Exploring differences gives rise
to negative as well as positive feelings. The challenge
in this phase is to avoid going back to the Buildup and
the Climax phases by keeping the overall emotional
tone positive while allowing negative feelings to grow
up and die down, but never take over.
Even after Germanys unconditional surrender in World
War II, it was necessary for the Allies to sit down and
negotiate detailed issues of governance with German
officials. After receiving Japans surrender, British
officials in some former colonies had to negotiate the
continued temporary employment of Japanese troops

Chapter 2


as a police force. The overall solution, unconditional

surrender, did not determine all these details. Nor did
it determine what the implicit fallback positions of
German or Japanese officials were in these
negotiations. It must have been non-cooperation in
various degrees. What if this non-cooperation had
been multiplied and increased by general failure to
negotiate satisfactory solutions? This is what must
have happened in the negotiations between German
invading forces and the at-first-welcoming inhabitants
of the Ukraine and other areas of the Soviet Union.
We suppose the result was first a return to the Buildup
phase where the position taken was Treat us better.
In the immediately ensuing Climax phase, the
Ukrainians came to a realization that this position
would be rejected. Finally, the Ukrainians reached a
Conflict phase where they decided to continue the war
with armed resistance.
Note that discussions at the Resolution phase need
not solve all the problems raised. By recursion, that
would mean all details had to be solved before
outline agreement could be reached, which is absurd
(e.g., agreeing the U.S. Constitution did not require
settling all political differences between parties). For
many problems, discussions will result in mere
agreement on parameters for future interactions. The
Resolution phase looks at details solely to confirm
the agreement reached within the simplistic, overall,
common reference frame arrived at following the last
Climax phase.


Confrontation Analysis

Modeling Considerations
Each difference or group of differences emerging in
the detailed negotiations of the Resolution phase can
be modeled by a submodel, where differences are
represented by differing positions. These submodels
then can be linked.
At one extreme, each separate issue would be
modeled separately. At the other extreme, all
differences would be brought together in a large cardtable. (Note that there are no practical limits as to how
many cards you can put in a card-table model.) The
chosen modeling procedure should reflect a decision
(which itself can be guided by building different models)
as to whether to divide issues into separate groups to
solve separately, or to solve them together. Solving
issues together (i.e., modeling them in one card-table)
allows consideration of trade-offs between issues.
Modeling them in separate but linked card-tables does
not allow consideration of trade-offs (i.e., it is difficult
to model an agreement that consists of a quid pro
quo between the issues represented in one card-table
and those represented in another). Separate but linked
consideration does allow the effect of issues on one
another, inasmuch as reaching a certain agreement
in one card-table may have an effect, through causal
linkages, on players preferences and the cards they
are able to play in another, linked card-table.

Phase 6: Implementation
We have seen several possible kinds of
Implementation of what is decided in a conflict
resolution process. What is implemented may include
any of the following:

Chapter 2


Resolution of the confrontationWhat is

implemented is the common position of all parties
in their common reference frame, with details
having been confirmed in the Resolution phase
(e.g., the overall agreement between parties in
Northern Ireland reached during Easter 1998)
False resolutionA common, agreed position
exists but is not implemented because parties try
to deceive others (e.g., the end of the Nazi
Soviet pact in 1941)
ConflictImplementation of the threatened
future, where all parties decide in the Conflict
phase to implement their fallback positions (e.g.,
Britains 1939 declaration of war on Germany)
Flunked conflictResolution has not been
achieved, and the parties declare the threatened
future should be enacted; however, it is not
enacted because one or more of the parties
decides not to implement its fallback position
(e.g., the 1993 aftermath of Serbian rejection of
the VanceOwen agreement).
These are four ways a confrontation may end. A fifth
way is by interruption (i.e., by a breach of the
informational closure that stabilizes the assumptions
underlying the confrontation). If the parties receive
new, relevant information from exogenous sources that
changes their assumptions and expectations as
created in the Scene-setting phase, then all bets are
off. If the parties to the present confrontation continue
to interact, it will be through conducting or resolving a
new confrontation, one that starts with a new scene
being set (e.g., a new confrontation started when the


Confrontation Analysis

Japanese attack on Pearl Harbor changed the

assumptions underlying the confrontation between the
United States and British parties as to whether the
United States should enter the war. The confrontation
was speedily resolved in favor of United States entry
into the war).

Implementation and the Unexpected

Interruption, by definition, creates an unexpected
future. The same generally is true of the other four
ways a confrontation may end. These other four
endings attempt to implement a known future, but this
attempt generally fails because it contains surprises.
We cannot foresee the future.
This is more than a matter of uncertainty (a known set
of possibilities with known probabilities attached to
them) or risk (a known set of possibilities with unknown
probabilities). Uncertainty and risk apply to the future
as projected by parties beforehand. The future that is
actually realized is usually one the parties did not think
of as a possibility.
The lack of foreseeable future is well known to warriors.
As Clausewitz emphasizes, parties projections of what
will happen if a threatened future is one of armed
violence usually fall wide of the mark. The point for
confrontation analysis is that these projections, not
the reality, are what persuade players to choose one
path or another.

Chapter 2


Unexpected Contingencies and the Need for

Positive Feelings
Agreements between parties are necessarily
contingent on the assumptions about the future the
parties made at the time, even if they know these
assumptions will turn out wrong. They know their
agreement will have to be reinterpreted by each of
them under circumstances not foreseen in the
agreement. This means that each agreement involving
future cooperation between parties needs to contain
a general card for each one to interpret this agreement
with the others interests in mind. To make the playing
of this card credible, a general cementing of feelings
of long-term goodwill or love is needed between
parties. Such feelings are well known to warriors. A
platoon going into action needs each member to
credibly play the card, Ill risk my life to save yours,
whatever happens; hence the love between soldiers.
Similarly, cooperation between different national
contingents in a combined command requires the
buildup of positive feelings.

Reversible and
Irreversible Decisions
Every event is irreversible, in one sense. If it happened,
it happened. Yet the logic of the conflict resolution
process as shown in figure 2 denies this. To resolve
or try to resolve our differences requires a time and
space where the only things that happen are like the
putting down and picking up of cards. They are seen
as reversible. We need to be able to say, They said
that, but we can change their mind. We can reverse
what was said. This time and space, created in the


Confrontation Analysis

Scene-setting phase by informational closure, is a

place where everything is reversible because
everything is seen primarily as a message, symbolic
of what we intend to do, rather than something done
for its own sake. The logic follows this thought: Nothing
is irreversible until the end (i.e., until one of the five
This means that until the Implementation phase, or
until the conflict resolution process is interrupted, all
that can happen is communication of intentions,
beliefs, values, and reasons.
Holbrooke walked out on Milosevic and Mladic at their
meeting in September 1995, thereby taking the
confrontation to the Conflict phase, but he was able
to resume talks with them when they relented, thus
moving back again to the Climax and from there to a
new Buildup.
Because nothing irreversible can happen during the
first five phases, any cards whose playing is by nature
irreversible cannot be played until the Implementation
phase. This rule is merely a matter of definition. It
means that a significantly irreversible decision would
be taken to represent the beginning of a new
confrontation, different from the old one inasmuch as
that decision would now be irreversibly fixed. By the
same token, if a card is by nature reversible, then an
important, even necessary way of signaling an
intention to implement it may be to start doing so.
The NATO air strikes that accompanied the same
HolbrookeMilosevicMladic meeting had, of course,
irreversible effects in damaging Serbian equipment
and taking Serbian lives; but their real function was to

Chapter 2


signal NATOs conditional intention to continue

attacking the Serbs. This is shown by the need to let
the Serb leadership know the strikes were going on,
and why, even though by the principle of surprise
(Alberts and Hayes, 1995, pp. 29-30), the attacks
would have been more effective in destroying the
Serbs war-fighting capacity if they had been unaware
of the strikes or the intention behind them.
The air strikes function as a signal depended on their
being reversible in the large (i.e., on the possibility of
stopping them if the Serbs relented). Of course, if
continued long enough, the air strikes might have
caused so much damage as to create an irreversibly
changed situation, a fact that set a kind of deadline
for the confrontation to be resolved. Yet the air strikes
were, by nature, sufficiently reversible that the signal,
Well bomb you, had to be, at a certain point,
reinforced by starting to do it or it would not have
been credible. The point at which this need for
credibility kicked in was partly set by the Wests
previous undermining of its own credibility.
By contrast, the Cold War nuclear threat, Well take
out Moscow, could not be and was not expected to
demonstrate its credibility by being carried out. It was
too irreversible.

You Can Destroy a Parking Lot Only Once

There is only one Moscow; hence by carrying out a
threat to nuke Moscow, you lose a card from your hand.
You have one less means of exercising pressure.
As an example, there is a possibly apocryphal story
that tells of a Bosnian Serb leader who was opposed


Confrontation Analysis

to a cease-fire receiving a phone call from NATO.

Look out the window. See your parking lot? With your
Range Rover in it? That noise is a British jet. Itll be
with you in two minutes. Know whats in its sights?
Your parking lot. Now perhaps youd like to change
your mind The General? He is standing beside me.
The Bosnian Serb changed his mind. The point of the
story is that if the parking lot had actually been
destroyed, such pressure could no longer be placed
on the Bosnian Serb leader. Destroy his house, his
family, and everything he values, and he has nothing
left to lose. Consequently it may be impossible to put
any pressure on him to concede.
This contrast between destroying other-party assets
to win a battle and to win a confrontation could not be
starker. Each asset I destroy increases my relative
physical strength but loses me a bargaining card;
therefore, it is counter-productive to destroy other party
assets when the following situations exist:
I have sufficient preponderance of physical
force not to need the physical strengthening I
get from destroying his assets. In general, the
United States and its allies have this kind of
preponderance in the post-Cold War world,
where we are typically the strong fighting the
weak; nevertheless, guaranteeing the physical
security of our forces may require selective
destruction of assets.
I dont need to start destroying his assets to
make it credible that Im prepared to. This
refers to the message-sending function of asset
destruction. The problem I may face is the need
to use up some of my bargaining assets to make

Chapter 2


the rest credible. This problem may be solved by

minimizing actual destruction while maximizing
its psychological effect and, at the same time,
sending the right messages to accompany it.

War-Fighting That Is Not

We should stress that the discussion in this chapter
has concerned a military campaign whose purpose is
to conduct a confrontation and, if possible, resolve it
on our terms. In this context, the primary purpose of
much of the military action that takes place is to send
a message; however, this describes only one kind of
campaign, the kind usually called an OOTW, although
the threat of war normally forms part of it, at least in
the case of peace operations. Peace is kept by credibly
threatening war. This is why such missions are
undertaken by the military.
In chapter 1, we argued that such missions are
increasingly common in the post-Cold War world;
however, military missions of the traditional,
Clausewitzian kind also occur. Here the mission is
simply and directly to destroy the military assets of
the other side, thereby lessening or eliminating their
ability to do us harm. Message-sending has little to
do with it.
The purpose of Clausewitzian war-fighting is, in
general, to set the scene for a subsequent peace in
which the victor can dictate terms, having achieved a
monopoly of military force. As Clausewitz says, while
stressing that war is always a means to a political end:
Whatever may take place subsequently, we must


Confrontation Analysis

always look upon the object as attained, and the

business of war as ended, by a peace. Clausewitzian
war is undertaken, not primarily to send a message,
but to change the facts on the ground so that
subsequent conflict-resolution processes may have a
more favorable outcome.
We can fit direct, purely military action into our model.
It can be seen as belonging to the Scene-setting phase
of figure 2. War of this kind is launched by politicians
who foresee that if they and others enter negotiations
equipped with the cards they presently hold, the result
will be unfavorable to them; therefore, they act to take
away certain cards held by others, cards consisting of
effective military response. Surprise and other
generally accepted principles of war apply fully.
Preemptive action of a traditional military kind also
may be taken to interrupt a process of conflict
resolution to be able to start another on more favorable
terms, after others have been deprived of certain
cards. This may happen in two following ways:
The aggressor may intend from the beginning to
take the preemptive action. It may simply use the
process of confrontation and conflict resolution
to lull others into lowering their defenses.
Alternatively, the aggressors participation in
conflict resolution may be genuine up to a
certain point. It then receives new, exogenous
information, the result of which is to raise its
demands above what it can reasonably expect
others to accept, given the present distribution of
cards among the players. So it decides to
forcibly take certain cards from others hands.

Chapter 2


Hitlers decision to interrupt discussions with the

Soviet Union and invade Russia may be an example
of the latter kind, looked at from a short-term
viewpoint. In a longer-term perspective, it was
probably a case of the first kind. He always intended
to attack the Soviets eventually.
Another example might occur if Iraq were discovered
to possess a nuclear bomb, and the Allies knew its
location. To prevent him threatening the West with it,
the decision might be made to destroy the weapon by
a commando or air raid. If successful, this would
remove the card, Use nuclear weapon, from
Saddams hand.
These examples are Clausewitzian in that war is being
waged for its physical effect in removing anothers
ability to harm us. Clausewitzian war may, of course,
also be waged by a defender of the peace in reaction
to an act of aggression.
We are not seeking to say anything original about
traditional warfighting, merely to contrast it with OOTW
and to point out the essential difference. Whereas
traditional war-fighting merely precedes and sets the
scene for the message-exchanging activity of conflict
resolution, an OOTW consists in large part of this
message-exchanging activity itself. It consists of
conducting a sequence of confrontations and
attempting to resolve them in such a way as to resolve
the issues in accordance with our objectives. That is
why the primary function of most actions taken in
OOTW is to send a message.


Confrontation Analysis

Summary of Chapter 2
An example of a UN force dealing with roadblocks
showed how confrontations at different command
levels are linked together. Handling them in a
coordinated way reinforces the effectiveness of
messages sent, and we point out that handling
confrontations is essentially a matter of messagesending, not physical operations.
Conflict resolution is a natural human process divisible
into six stages: Scene-setting (when the problem is
posed within a given context); Buildup (when players
take positions and fallback positions within a common
reference frame); Climax (when they must either
redefine their positions and their frame, or else fall
into conflict); Resolution (by having changed they find
they agree and can trust each other to implement their
agreement); Conflict (when having failed to agree, they
are faced with having to carry out their fallback
positions); and Implementation (when they carry out
or decide not to carry out either the agreement they
have reached or the conflict they have fallen into).
A common reference frame embodies the minimal set
of assumptions players need to share to communicate.
It needs to be simple, it may be deceptive (because
players can deceive each other), and it is interpreted
differently by players with differing values.
At the Climax, players use emotion and reason
(positive and negative) to change themselves and one
another. Reason and emotion are each relatively
ineffective without the other. Arguments based on the
players common interests are most effective;
consequently, players conducting confrontations tend

Chapter 2


to build up the preferences of a superplayer to which

they all belong.
Conflict may be flunked (i.e., not implemented because
of unwillingness to carry out threats that have been
made), just as agreements may be betrayed. Because
the future cannot be foreseen, conflict outcomes in
any case rarely turn out as projected, and agreements
need goodwill between participants because they will
need to be maintained in circumstances not foreseen
when they were launched.
During conflict resolution, nothing irreversible can be
done, because all that is supposed to happen (until
the Implementation phase, provided the process is not
interrupted) is dialogue. Credibility may require
beginning to carry out threats, but that is still regarded
as a way of sending a message. There is a problem
when we send a threatening message by destroying
assets. This defeats the purpose of the message
because the threat does not exist when the assets
have been destroyed; therefore, it may be best to
destroy low-value or replaceable assets provided this
has sufficient psychological impact.

Chapter 3
How Parties Pressure
Each Other:
Cooperation and Trust

ur continuing aim is to show how a commander

can form a strategy to win a confrontation in the
sense of getting his position genuinely accepted.
Of course, the most excellent strategy conceivable
cannot by itself guarantee victory, any more than it
can in battle-fighting. Superior strength and morale
are also important, just as they are in battle-fighting,
although in a different way. There is a further
complication in that mission objectives themselves
may change in the course of a confrontation, and with
them the very definition of success; nevertheless, it is
important, as in warfighting, to have the best possible
strategy for obtaining given objectives. How is this
In the confrontation shown in table 2, how should the
commander work out a strategy for getting the militia
to comply with the agreement to cease ethnic
cleansing and allow free movement? How should he
devolve his strategy to other levels of his command,
such as the level of the commander in table 3?
To answer these questions, one part of the resolution
process is of prime importance. It is the Climax phase,



Confrontation Analysis

as shown in figure 2, where changes take place in

characters attitudes. In this chapter we analyze these
changes and show how they are brought about.
Understanding them is the key to success. It must be,
because the commander requires these changes to
go in the direction of his objectives.
As we discuss below, the commanders job as a peaceoperations professional is to build on the fact that there
is a tendency for changes to occur unless and until a
full resolution is reached (i.e., until all parties have
converged on a single solution they can trust each
other to implement). He must direct and orchestrate
these tendencies to ensure that resolution is achieved
as close as possible to his position.

Games vs. Drama

Therefore we take up the question, What drives the
changes in attitudes, beliefs, and objectives that take
place in the Climax phase shown in figure 2?
To lay the foundations for an answer, we will look at
the main assumptions underlying von Neumanns
theory of games and the more general approach within
the social sciences that calls itself rational choice
theory and reaches its fullest development in game
theory. (For game theory, see von Neumann and
Morgenstern, 1959. For a clear account of its present
reformulation, following decades of development, see
Osborne and Rubenstein, 1994.)
Our actual question is debarred by game theory. The
subject defines itself in such a way that the question,
What drives these changes? cannot be asked. It

Chapter 3


cannot because rational choice theory has a mission.

It tries to take the idea as far as it can that all choice is
rational; and the term rational is used to mean
pursuing fixed, given preferences within a fixed, given
frame. To assume rationality in this sense is to
assume that the kind of change depicted in figure 2 is
impossible. Characters beliefs about the frame and
their preferences for possibilities within it (although
not their positions) are assumed to be fixed. Of course,
exogenous changes in these beliefs and preferences
(interruptions) brought about by new information
coming from outside, are allowed, just as we allow
them. But change brought about merely by characters
pressuring each other is assumed not to occur.
Why do we discuss this rationalistic view taken by
game theory if it excludes the possibility of answering
our question? One reason is that the reader may have
come across it. It is important throughout the social
sciences, it dominates our understanding of
economics, and it is influential in political science,
negotiation theory, and military theory.
That is not our main reason for discussing it. The main
reason is that in pursuing its limited definition of
rationality (the pursuit of fixed preferences within a
fixed frame) game theory uncovers numerous
dilemmas, the best-known being prisoners dilemma
(see in particular the early discussions by Schelling,
1960; Rapoport, 1964, 1966; Howard, 1971). These
dilemmas show in various ways that the pursuit of fixed
preferences within a fixed frame is not what it seems;
to achieve outcomes high on ones list of preferences,
it may actually be advantageous to forego pursuing
ones preferences because characters generally need


Confrontation Analysis

to make threats and promises they would prefer not

to carry out.
The West wanted the Serbs to stop invading
Bosnia. To achieve this, the West had to threaten
to do something it preferred not to do, attack the
Serbs. Fixed, known pursuit of the latter
preference (not to attack) prevented satisfaction
of the former (Serb cessation of violence).
The Serbs did not want to be bombed. They also
wanted to take over much of Bosnia. When the
West made its ultimatum credible, Serbian
pursuit of its preference for taking over Bosnia
prevented it from satisfying its first, greater
preference, cessation of bombing.
Game-theoretic dilemmas are created by a tension or
contradiction between the values a character wants
to pursue and the credibility it (the character) needs
to have to pursue those preferences. The dilemmas
so created are dilemmas not just for the theorist, but
for the players themselves. They are in paradoxical
situations: the prisoners in prisoners dilemma actually
see that if each chooses so as to guarantee the worst
possible outcome for itself (given the others choice),
both do better for themselves than if each guarantees
the best possible outcome for itself (again, given the
others choice). (Howard, 1971, p. 45).
This illustrates that game-theoretic dilemmas bring
players face to face with psychologically intolerable
paradoxes and gives us a clue to explaining the
changes that take place in the Climax phase of a
confrontation. We make the following hypothesis:
players make changes in their common reference

Chapter 3


frame, their own and others preferences, and the

purportedly final positions they have taken up in an
attempt to eliminate paradoxical dilemmas. We
hypothesize, therefore, that emotions are aroused in
a predictable way by game-theoretic dilemmas. A
player is pointed in the direction of dilemma-eliminating
changes. Reason and evidence are invoked to justify,
if possible, changes that emotion has pointed to.
For example, Western public opinion eventually
became sufficiently emotionally aroused against the
Serbs to eliminate the dilemma the West faced, that
we preferred not to attack the Serbs even though we
needed to make credible our threat to do so. Our antiSerb feelings, aroused by this dilemma, were
rationalized by demonizing the Serbs (i.e., by
propagating the idea that they were evil and supporting
it by selectively highlighting the atrocities they
committed while ignoring those committed against
them). After we achieved such rationalization, our
preferences changed. We now preferred to attack the
Serbs rather than let them continue attacking Bosnia.
The proposition that emotion evokes the use of reason
and evidence to try to bring about dilemma-eliminating
change is the basis of the discipline, drama theory
that underpins confrontation analysis. (See Howard,
Bennett, Bryant, and Bradley, 1992; Howard, 1994,
1994a, 1996, 1998; Bennett and Howard, 1996;
Bryant, 1997. See also Nigel Howard Systems, 1992
It is, perhaps, a disconcerting proposition if one
believes that reason and evidence should be used to
try to reach objective, unbiased conclusions rather than
to prove a point given beforehand. This belief is not


Confrontation Analysis

generally well founded. We find the very epitome of

objective reasoning in scientific method, the closest
method for arriving at objective truth. Even so, students
of scientific method stress that scientists too must have
a hypothesis to prove or disprove before selecting
evidence or using reason (see Popper, 1959; Kuhn,
1962). The important thing about the collocation of
reason and evidence is not that it is impartial or
disinterestedin the sense of having nothing to prove.
That is not the case. Its importance is that, if effectively
presented, it compels belief in what it aims to prove.
That is precisely its function in a drama-theoretic
confrontation. Emotion may make me heard; I need
reason and evidence to make me believed. The
reason I need to be believed is that I need to make
my position credible.

Why the Drama Metaphor?

The word drama replaces the word game in the
theory that underlies our approach. Practical users
need not bother with fundamental theory;
nevertheless, they may ask why. What does the term
drama point to that game obscures? Both games
and drama are role-playing, leisure activities that are
used to illuminate life itself. The game metaphor was
used by the military for training purposes long before
game theory was conceived. Do we need a change of
metaphor to understand the basic nature of OOTW?
To see that we do, compare behavior in game-playing
with behavior in drama. A game lays out a fixed set of
choices (sequential or simultaneous), specifies what
outcomes (stochastic or deterministic) to expect from
each combination of choices, and requires players to

Chapter 3


have specific preferences over outcomes. It is

forbidden either to vary your preferences (e.g., prefer
another to win, or to invent other choices, such as
move anothers pieces). The result is that as long as
actors play the game, they have nothing to do but
predict what others will choose and choose the mostpreferred outcome for themselves, given their
predictions. This is precisely what game theorists
mean by rationality.
By contrast, in a drama we see characters being
emotional and irrational and participating in rational
debate. In doing so, they pass through a crucible that
changes them. Both their value systems and their views
of reality change. The very rules of the game change.
The chief interest of drama is in seeing how this
happens. The main thing we learn from drama is not,
as with games, how to be rational in the sense of
pursuing fixed preferences against others with fixed
preferences, all the time keeping to fixed assumptions
about what is possible. Instead, we learn how players
(now called characters) in interaction with each other
use reason (in another sense of the term) and emotion
to change their common assumptions and to work on
the value roots of their own and others preferences
to change them also.
The way a drama ends differs from the ending of a
game. A game ends with the victory of one side. Losers
are left in an artificial state of frustrated discontent. It
is artificial in the sense that only the arbitrary rules of
the game prevent the losers from doing anything about
it. If they choose to operate outside these rules, they
may resort to throwing the pieces on the floor, wrecking
the grounds, or attacking their opponents.


Confrontation Analysis

By contrast, a drama ends when none of the characters

has anything left to hope for or to fear. This
characterizes a state of stability, with no further
tendency for characters expectations to change,
although new external information (interruptions) may
disrupt this state. In a real-world drama, the
dnouement never turns out as expected, a fact
passed over in fiction by not showing what happens
afterwards as a result of the agreements and
understandings reached.
The end of a drama is nevertheless characterized by
complete stability of expectations, which may be
brought about in two different ways represented by
tragic and happy endings. A tragic ending is stable
because characters fears have been realized and their
hopes destroyed; a happy ending is stable because
characters hopes are realized and fears banished. In
either case, they no longer have hopes or fears.
From a military commanders viewpoint, what matters
is that the ending of a drama represents stability, while
the ending of a game does not. Losers of a game look
forward to another round, if they do not demand a
replay or resort to violence. Characters at the end of a
drama are content, seeing no alternative to acceptance
of what they have.
Such contentment, enforced or willing, must be the
commanders object in an OOTW. The parties need
to be brought into a state of contentment with a future
that consists of fulfillment of the commanders mission.
They may, of course, be in various states of discontent
and conflict in regard to other matters, particularly
details. The aim of a peace-support operation is
generally to bring about a broad degree of peace and

Chapter 3


order, not to settle every issue between inhabitants of

a region; nevertheless, in regard to the broad societal
issues that he has to settle, the commanders object
must be to make the parties content with his position.
The means for obtaining that kind of stability through
contentment are examined in drama theory.

The Need for Game Theory

We must stress that despite its willingness to violate
game theorys basic assumptions, drama theory needs
game theory. In some places it needs game theory
directly. When for any reason characters cannot
interact with each other to negotiate a resolution (i.e.,
when preplay communication is disallowed) drama
theory reduces to game theory.
This, in particular, is the state of affairs when parties
enter the Conflict phase as shown in figure 2. At this
point they have failed to reach agreement and must
decide whether to implement the fallback positions
they are committed to implementing or, if not, what
else to do. They must make this decision having cut
off communications with each otherunless, that is,
they decide to go back to the Climax phase by asking
to reopen negotiations on the ground that they have a
change to announce. Failing this, the calculations they
must make include hypothesizing if the others will stick
to their fallback positions. If so, what should I do? If
not, what else will they do? What will they think I am
going to do? All these responses are game-theoretic.
This is true only of interactions between the characters
as such. Internal confrontations between the
subcharacters making up a character generally


Confrontation Analysis

continue to be drama-theoretic. Even so, it is clearly

one way in which game theory is part of drama theory.
Warfighting itself is essentially game-theoretic. In the
traditional Clausewitzian model, the military takes over
when politicians fail to resolve a confrontation in which
Conflict (i.e., the Conflict phase as shown in figure 2)
means war. In such case, the development of a military
strategy to destroy the enemys war-fighting capacity
is essentially an application of game theory, regardless
of whether quantitative game theory (the form in which
it is usually presented) is considered useful. On the
other hand, internal relations between the different
units of a joint or combined force are drama-theoretic.

Dependence on Game-Theoretic Dilemmas

A fundamental reason drama theory needs game
theory is that it uses the dilemmas produced by game
theory to analyze, predict, and understand the
pressures on characters in a Climax phase to change
their positions, preferences, and common reference
frame. The following logical derivation of drama theory
shows how.
Initially we suppose that characters see their
common reference frame as fixed, and hence
see themselves in a game. This is because they
have assumed this frame for purposes of
communication and for the time being, cease to
question it.
As the characters see themselves in a game,
they try to behave rationally in the gametheoretic sense. This brings them up against
game-theoretic dilemmas.

Chapter 3


The dilemmas then cause the characters to feel

positive and negative emotions. The emotions
felt depend on the dilemmas the characters face
and the ways they deal with them.
Emotions may cause them to behave irrationally
(i.e., not optimally), change their preferences, or
search for new cards to play or new characters
to introduce. By these means the dilemmas may
be eliminated.
As an alternative, the characters may solve their
dilemmas by changing their positions.
A general alternative to these methods of
dilemma-elimination is deception. Instead of
becoming irrational or changing preferences,
cards, or positions, characters may pretend to do
so. Note that deceitful persuasion changes the
common reference frame just as much as nondeceitful persuasion.
What makes deceit attractive for one character
creates disbelief in another. Such disbelief
must be overcome to effect the desired change
in the common reference frame. To overcome
it, characters construct logical arguments,
show evidence, and appeal to generally
accepted standards.
Characters arguments cannot be value-free; to
make sense, arguments generally must assume
the pursuit of some common interests or
objectives. If objectives are not shared,
arguments based on them will seem insincere or
unappealing. Characters are thus led to construct
rational arguments in the common interest.


Confrontation Analysis

Successful arguments of this kind have the effect

of building up the preferences and attitudes of a
supercharacter formed by an alliance of the
characters. Individual characters preferences do
not become the same as the supercharacters,
but become such that solving subcharacter
confrontations becomes a mechanism causing
the supercharacter to function as a character.
The supercharacter generally will be a character
in a larger drama, in which it too tries to behave
rationally and so confronts dilemmas.
Starting from the assumption that characters try to be
rational within a fixed frame, drama theory shows how
they are led to behave irrationally and change the
frame, so creating the possibility of rational behavior
at a higher level. This is the deep sense in which drama
theory depends on game theory.

The Cooperation Dilemma

Having seen in general how the encounter with gametheoretic dilemmas leads players to change, we look
in detail at how this happens.
There are six dilemmas that can place pressure on
characters: cooperation, trust, deterrence, inducement,
threat, and positioning. We will examine each and
discuss the kinds of change necessary to overcome it.
Each dilemma is illustrated with examples, including a
reference to the roadblocks-removal problem in tables
2 and 3.
The cooperation dilemma faces a character that is
tempted to defect from its own position. (This arises

Chapter 3


because my position generally will contain things I do

not want, but have included as concessions to others.
I may be tempted not to carry them out).
A cooperation dilemma faces a character at a card
table when, by changing its own selection of cards,
the character can move from its own position to a future
that it likes just as well or better. Table 2, column P,
illustrates. Here, the ethnic militia has accepted the
commanders position, thus also adopting that position
for itself. Actually, the militia prefers the future (not
shown on the table), otherwise the same as P, in which
it does not play the card, Allow free movement. The
militia can move to this future (if the commander carries
out his part of P) merely by changing its own selection
of cards; therefore, the militias own preference gives
it a cooperation dilemma.
The cooperation dilemma can be thought of as the
drunks dilemma. Imagine an alcoholic whose wife
threatens to leave him if he will not stop drinking. He
does not want her to leave, so he swears he will stop.
He knows, and knows that she knows, that if she stays
he will not be able to keep his promise, but will contine
to drink.

General Statement of the Cooperation Dilemma

Generally stated, the cooperation dilemma faces a
character that belongs to a group (subset) of characters,
all of whom can, by changing just their own card
selections, move from the characters position to
another future they all like just as well or better.
For example, assume the drunk depends on a visiting
friend for his supplies of booze, as shown in table 4.


Confrontation Analysis

Table 4. The Drunk, His Wife, and His Friend.

Chapter 3


The friends position, F, is that the husband should be

allowed to drink (the friend likes someone to drink with)
and the wife should let him have the money to do so
(rather than forcing him to rely on the friend). The
husband (the drunk) still swears to his wife that he will
quit; thus the drunk and the wife share the position P.
The drunk accepts P because he fears the threatened
future t; however, he knows and she knows that he
will be tempted to get his friend to assist him in
defecting from position P: both the friend and the drunk
prefer column d over position P. They can get to d
from position P just by changing their own joint card
selection, assuming the wife does not change hers.
Note that the wife may not be able to detect a move
from P to d. Because d is the default future (i.e., the
future they are presently in, and will continue to be in
if present policies continue to be pursued), the wife
suspects that despite his promises, he will secretly
carry on drinking, supplied by his friend.

Eliminating the Cooperation Dilemma

A character that faces a cooperation dilemma must, if
it wants itself and others to genuinely accept its
position, make them believe an incredible promise.
How can the character do this?
The character may abandon its position. For example,
the ethnic militia may openly say it does not agree to
free movement, or the drunk may tell his wife he will
not give up drinking. In each case (although not in all
cases), this lands the character in other dilemmas;
however, it eliminates the cooperation dilemma.
If a character does not abandon its position, it must
try to make its promise credible.


Confrontation Analysis

What emotion then moves the character? In general,

positive emotion, arising from the urge to reassure,
will cause the character to feel and show goodwill,
friendship, or love (the appropriate term depends on
the types of characters and relationships involved). If
the character seems sulky or reluctant or aggressive,
its promises will not easily be believed.
Positive rationalizations go with positive emotion.
The husband, in an attempt to be convincing, may
explain how and why he has decided to change his
life. The militia may give reasons why it has changed
its mind and decided to agree with free movement.
Unless reasons are given, the characters conversion
will be unconvincing.
Emotion and reason are to some extent substitutes.
Strong emotion without adequate reasons may
convince others that the character can be trusted,
temporarily at least. They may think, He feels this
strongly enough to go against his own interests or
desires (irrationally, in the game-theoretic sense of
rationality). On the other hand, strong reasons may
convince others of the characters trustworthiness,
even if the reasons are accompanied by the wrong
emotional signals. Emotion and reason are most
effective when they accompany and support each
other. Then the characters emotion shows that its
preferences are in the process of changing (so that in
the game-theoretic sense, the character must be
temporarily irrational, while moving between
preferences); the characters reasons give cause to
believe the change will hold.
Finally, a character may take irreversible actions to
eliminate the cooperation dilemma by removing its own

Chapter 3


temptation to defect. For example, the drunk may tell

his accommodating friend to leave the house. The
commander of the ethnic militia may make public
broadcasts threatening discipline against those who
set up roadblocks, thereby irreversibly tying his
reputation as a leader to their removal and making
himself not prefer not to take action against them.

Deceit, Disbelief, and Rational Arguments in the

Common Interest
Is a character that projects emotions and gives reasons
in the manner suggested above deceitful? Possibly,
but not necessarily.
Irreversible actions actually may change preferences,
as in the examples given. In addition, attitudes, values,
preferences, and beliefs can change for emotionalrational reasons. We propose that it is just these forces
of emotion and rationalization that cause genuine
change. In convincing others, you may convince
yourself. Indeed, if genuine change were impossible,
deceit would be impossible. It generally is not possible
to make others believe the impossible.
However, it is important that many of the above
methods of eliminating the cooperation dilemma may
indeed be carried out with deceit (i.e., not carried out
in actuality, only in pretense). Deception means that
the dilemma is eliminated from the characters
common reference frame (their common view of the
situation, needed for communication). That is all that
dilemma elimination requires. Any deceit, if successful,
will not be discovered, at least until the Implementation
phase, as shown in figure 2. Whether it is discovered


Confrontation Analysis

then, later, or ever depends on characters information

at later stages.
Now any situation in which deceit is profitable is one
where disbelief is appropriate. Any of the dilemmaelimination methods described above may be
counter-productive and induce disbelief: They want
me to believe this; maybe I shouldnt. Hamlet says
on such an occasion: Methinks the lady doth protest
too much.
We derive a theorem: No one should ever believe
anyone, because if you tell me something I can deduce
that you want me to believe it, which gives me a reason
not to, since presumably you would want me to believe
it whether it were true or not (since in either case you
would succeed in making it part of our common
reference frame).
To rescue us from this conceptual abyss, recall that it
is the characteristic of reason and evidence to compel
belief. They have this characteristic because they
make deceit difficult or impossible. Of course, difficult
is not impossible. Reason and evidence may require
exhaustive scrutiny to compel belief. However, it is
their essential characteristic that they can do so
because they can uncover deceit.
This, then, is why reason and evidence are needed to
back up the more primitive method of emotion in
attaining credibility. They are needed to overcome
disbelief. We may speculate that humans developed
rational argument in the course of evolution simply
because they developed the capacity to deceive. We
may reason as follows: After humans began to find
an evolutionary niche in their enhanced ability to

Chapter 3


cooperate, deception became an advantageous

strategy for an individual or subgroup by enabling them
to share in the payoff from cooperation without paying
their dues; in the presence of deception, disbelief
became advantageous; and given disbelief, the ability
to compel belief by rational argument became
advantageous in turn.
Rational argument requires an ultimate framework of
the common interest within which to operate, because
it is always possible to go behind any argument or
evidence and ask, Why? How? Where from? For this
basic reason, arguments must in the end be founded
on common values to be convincing. Common
interests will, however, always exist when there is need
to make a promise credible because such a need
implies a common preference for having the promise
believed and kept (as in column P) rather than not
believed (as in column t ). This common interest,
generalized, can be made a foundation for common
values and hence for acceptable arguments.
The ethnic militia, if it truly wants to convince the
commander of its sincerity, may draw upon a range
of common-interest arguments and a fund of possible
common values in favor of freedom of movement and
against setting up roadblocks. Similarly the drunk may
draw on the values and interests he and his wife have
in common in their marriage, all of which will be
sacrificed if he continues to drink. In each case, the
common-interest arguments made may need to be
backed up by evidence of the value change they
argue for.
In summary, we can derive from characters needs to
make their promises credible part of a process by which


Confrontation Analysis

they build up or maintain a conception of their common

interest and so become an effective single supercharacter
in larger superconfrontations. From other dilemmas, we
will derive other parts of this process.

Whether it is easy or hard to change attitudes, beliefs,
values, and preferences depends on the strength of
what must be changed. It is easy for me to prefer that
you precede me through a door, although note that I
feel an access of goodwill toward you even as I make
this slight preference change. It would be hard for me
to prefer to die to save you.
The many factors that make it hard to change a common
reference frame are called friction. They involve many
things. In general, friction arises when values are hard
to change and evidence hard to escape.
In this general sense, however, values may be more
changeable than the word suggests, and involve no more
than trivial preferences or marginally different ways of
applying deeply held value systems. For example, an
officer who values discipline may make many small
decisions in the course of applying it; these represent
adjustments to what he means by this overarching value,
and thus they are mini-value changes.

Why Do People Keep Their Promises?

Why is it we often can rely on people to keep a promise
to do things that they would prefer not to do if they
had not promised? Often it is because the friction they
find in changing their preferences between futures is
slight enough to be overcome by the emotional urge

Chapter 3


to come to an agreement and prefer to do as promised.

They feel an emotional, preference-changing,
dilemma-eliminating urge, however slight. To see this,
recall how you feel when promising to do something
you would prefer not to. You tend to feel and project
goodwill, provided you mean to keep the promise. That
projection of emotion tells others you can be trusted.
If it is not there, people would find it disturbing (provided
a preference change is indeed required; provided you
are promising something you would have preferred
not to do, rather than just notifying people of something
you would want to do anyway). The same projection
of positive emotion occurs when people change their
beliefs to reach an agreement (dilemma elimination
by belief-change).
In a further twist, strong, principled believers in keeping
their promises may show their trustworthiness by
projecting no emotion, signaling that having made a
promise, their principles infallibly kick in to make them
want to keep it. Such people lay themselves open to
being misunderstood by strangers.
When stakes are high, such as in matters of peace
and war, friction is too great for such social
mechanisms. Merely promising something will not
change preferences toward performing it.
The militia in table 2 wants to convince the commander
of its sincerity, to keep him from forcibly freeing up the
road network, but there is too much friction involved
(the militia wants to keep roadblocks in place as a
source of revenue, to keep out returning refugees, and
to control movements of groups that might threaten
the militia) for the militia to change its preferences and
genuinely prefer free movement. The militia cannot


Confrontation Analysis

help being insincere, despite wishing to convince the

commander of its sincerity.

Eliminating the Cooperation Dilemma by

Changing Position
So far in our discussion of ways of eliminating the
cooperation dilemma, we have assumed that a player
will maintain the aim of having its position accepted.
A players proclamation that it aims to achieve its
position may be deceptive. We have allowed for this.
The drunk or the ethnic militia secretly may intend to
defect, after having obtained acceptance of its position,
but a character practicing deception still wishes to have
its position accepted: it needs this to be able to defect
from it. In light of his wifes ultimatum, Stop or Ill leave
you, the drunk cannot carry on drinking until he has
made his wife believe he genuinely intends to stop. In
light of the commanders ultimatum, Stop or well use
force, the militia commander cannot carry on
obstructing roads without first convincing the
commander he intends to stop.
A dishonest negotiator is still a negotiator that aims to
have its position accepted.
By contrast, another way to escape the cooperation
dilemma is to actually change ones position (i.e., to
switch to a position that does not suffer from that
particular dilemma in that particular way, although it
may suffer from other dilemmas and even from another
cooperation dilemma).
What emotions accompany the abandonment of a
position? If a player has been determined to maintain
its position, sorrow and despair are feelings that help

Chapter 3


it to withdraw emotionally from the objectives it has

pursued, reassess the situation, and pursue less
ambitious aims. These emotions generally accompany
a committed players change of position.
In table 4, the husband would feel sorrow and despair
if his inability to stop drinking made him decide to
give up his wife (i.e., if he decided that the only way
to solve his cooperation dilemma at P was to shift
position by accepting the threatened future t,
consisting of his wife leaving him while he continues
to drink with his friend).
Alternatively, a player may welcome a change of
position. This is likely to be so with a player facing a
cooperation dilemma: Because I got myself into this
dilemma by accepting a position that is inferior to at
least one other possible position (the one I am tempted
to move to), I can probably think of other positions I
would prefer to go to, if I could get other players to
accept them. For example, there may be a position I
have relinquished unwillingly, under threat of
sanctions, but would like to go back to it if I can
persuade others to agree.
A player who for such reasons would welcome a change
of position may seize upon the dilemmas that beset its
present position as an opportunity to persuade others
that it might be a good idea to move away from it. We
can suppose, for example, that the ethnic militia in table
2 and the drunk in table 4 each have accepted position
P solely to escape t. This being so, they may use their
own cooperation dilemma at P as an argument for
abandoning or modifying this position. The drunk may
say, I know I agreed to stop drinking, but you know
how hard it will be for me. Would you really be able to


Confrontation Analysis

trust me? Maybe we should think of some compromise

arrangement I would actually be able to live with. The
militia commander may say, Agreed. Free movement.
But it will be tough to get my people to agree. I dont
know if I can do it. Is there a compromise position it
would be easier for me to sustain?
The emotions that accompany attempts to change
position in the latter case (when the aim is to move to
a preferred position, as distinct from one that is
accepted only because it is realistic) are, like despair
and sorrow, emotions of withdrawal. The emotional
tone here is skeptical, detached, and deflationary,
rather than despairing, because the character needs
to deflate emotions and puncture arguments that
suggest commitment to the present position by
pointing to other ways of solving its dilemmas.
If the wife or UN commander beams forth positive
emotion and constructive arguments as to how the
drunk may stop drinking or the militia commander
may free up the roads, the drunk or militia
commander, if they want to argue for a move to
another position, will try to deflate the emotion and
debunk the arguments by being skeptical, realistic,
and emotionally uninvolved with the problem. In this
way, they hope to initiate acceptance of another, more
acceptable position.

The Trust Dilemma

I face a trust dilemma when you are tempted to defect
from my position.
In card-table terms, a trust dilemma confronts me
when a group of one or more characters (not including

Chapter 3


me) can, by changing just their own card selections,

move from my position to another future they all like
just as well.
Note that when we all share a common position and I
have a trust dilemma, it must constitute a cooperation
dilemma for someone else.
In table 2, the commander has a trust dilemma
because he cannot trust the militia to allow free
movement. In table 4, the wifes trust dilemma is that
she cannot trust her husband not to continue drinking
alcohol supplied by his friend.
Is this the same dilemma stated twice? It is not,
because my inability to trust you is my dilemma; your
inability to be trustworthy is yours. There is a difference
between not being able to trust myself (thus being
potentially untrustworthy to others) and being unable
to trust others. Consider, for example, the case when
the ethnic militia in table 2 wants to escape its
cooperation dilemma by changing position, while the
UN commander wants to get rid of his trust dilemma
by persuading the ethnic militia to change its attitudes
or incentives. His trust dilemma is the militias
cooperation dilemma. Their perspectives are different.
When there is no common position, my trust dilemma
does not need to be someone elses cooperation
dilemma. In table 3, the local commander has a trust
dilemma because he is asking the local militia to agree
to something he cannot trust them to implement.
For example, if the militia agrees to remove the
roadblock (column C), it can move from C to its own
position, M, simply by taking back the Remove
roadblock card (the militia prefers M to C). The militia


Confrontation Analysis

can move unilaterally from the local commanders

position to a future it prefers. This gives the local
commander a trust dilemma. The lesson he can draw
from it is this: If the militia agrees, he will need to
make sure it does remove the roadblock, and does
not put it back when he goes away.
The same dilemma faced the United States when
President Richard Nixon and Secretary of State Henry
Kissinger attempted to negotiate with North Vietnam
to end U.S. involvement there. Put simply, the U.S.
position was: Well leave Vietnam if you agree not to
invade the South. North Vietnam replied: How can
we agree to such a thing? Would you believe us if we
did? The U.S. response was: Agree, damn you, or
well bomb you until you do. And the United States
did just that; it dropped more bombs on Hanoi than
had been dropped in the whole of World War II. The
North Vietnamese finally said: Okay, we agree. The
U.S. response was: We dont believe you. How could
we? Still, the United States left. North Vietnam then
invaded the South.
Clearly this account simplifies the negotiations;
nevertheless, it captures the essence of the United
States trust dilemma. In a moment of truth, negotiators
themselves simplify the issues to make sure they have
a common reference frame.

Eliminating the Trust Dilemma

As with the cooperation dilemma, I can eliminate my
trust dilemma by giving up my position. For example,
the local commander can cease to require the
roadblock to be removed. The wife can agree to let
the husband drink.

Chapter 3


The emotions that go with giving up a position one is

committed to are generally negative toward oneself
(e.g., despair, sorrow). Accompanying rationalizations
consist of finding reasons why the new position is not
so bad, after all, and may even be better than the one
abandoned. As with the cooperation dilemma, it is also
possible that the position being moved to is already
preferred, in which case the trust dilemma may be
welcomed as an argument in favor of changing position.
To maintain my position against a trust dilemma, I
must find ways to eliminate anothers temptation to
defect from it. The required emotional tone is then
positive and constructive, showing goodwill,
cooperativeness, and empathy with the other. This is
true even if I impose an automatic sanction to
eliminate the others temptation.
Suppose the local commander lets the local militia know
that, after the roadblock has been taken down, a
reconnaissance satellite will spot it if the roadblock goes
up again, and rocket fire automatically will be called
down. Despite the fact that violence is being threatened,
a positive, friendly tone will convey this information most
effectively, because its object is to elicit cooperation
with the commanders position. The commanders tone
should imply that violence will be unnecessary because
the militia will cooperate. Violence is there to stabilize
the agreement and provide a reason for the parties to
trust each other.
Rationalizations accompanying the projection of such
positive emotion also should be positive, aimed at
constructing a common position on the basis of
common interests. For example, the local commander,
even while giving notice of his automatic sanction,


Confrontation Analysis

should adduce all kinds of common-interest arguments

in favor of freedom of movement, such as rebuilding
the economy and bringing the country forward to join
the community of advanced nations. The aim is to bring
about willing compliance with his position. He should
assume this will be forthcoming and welcome it in a
spirit of goodwill.
The wife might let her husband know that she can
always spot his drinking and will leave without further
warning if he continues one more time. This is
automatic retaliation. At the same time, she should
build on their love for each other and their common
interest in the marriage succeeding provided he gives
up drinking.

Deceit, Disbelief, and Positive Argument

The local commanders threatened satellite
surveillance and automatic calling of fire may not exist
or be as efficient as he suggests; the wife may not be
able to spot her husbands drinking as infallibly as she
says; thus, these communications may be more or
less deceitful. Because this is so, the other side may
be more or less disbelieving. The fact that one side
has reason to produce belief, even if unjustified, makes
disbelief rational for the other side.
To combat disbelief, concrete evidence may be
produced that automatic retaliation is a reality.
Alongside or instead of this, positive arguments in favor
of the position advocated are a way to deflect disbelief.
A bald sanction invites disbelief and hence resistance,
leading to thoughts of how to get around it. Positive
reasoning on the basis of common interests directs
attention away from such reactions and toward seeing

Chapter 3


the threat of automatic retaliation as an excuse for

giving in to what, judged by such reasoning, seems
preferred on its own merits.
Thus rational common-interest arguments, positive
emotion, and evidence of automatic retaliation are, in a
way we have seen before, both possible substitutes for
each other and strongest when used together. If my
system for automatic retaliation is objectively
unconvincing, I may compensate for this by good
common-interest arguments, presented sympathetically.
At one extreme, I may be able to do without retaliation
altogether and change anothers mind merely by
common-interest arguments in favor of my position; vice
versa, I may compensate for weak arguments presented
badly by strong evidence of mechanisms for retaliation.

Modifications of Position
Your arguments will be most effective if you take
account of others attitudes and beliefs and pay
attention to their objections, incorporating them into
your position if possible.
In this way, their objections can be used to modify
your position itself. Assuming that your aim is to
maintain your position, not to change it, you may
nevertheless find that one way to safeguard the
essential features of a position may be to modify it to
make it more acceptable to others. This works as
follows. Building upon common interests naturally can
lead to the addition of cards that make our position
more favorable to the other, eliminating the dilemma
by making our position better for them than the
temptation to defect. Thus the wife or local commander
might promise the husband or local militia something


Confrontation Analysis

they have been wanting as a reward for going on the

wagon or removing the roadblock.
Such inducements will be more credible if stated in
terms of the common interest rather than in the
interests of those you are inducing, because commoninterest arguments give them reason to suppose you
will want to carry out the promise you are making.
They also will be more credible if first suggested by
themselves, because this makes the inducements
seem less likely to be a trick or trap.
History cannot be rewritten; nevertheless, it is
interesting to speculate how the United States might
possibly have eliminated its trust dilemma with North
Vietnam. Such speculation might include creating new
cards to put in characters hands: automatic retaliation
(U.S. forces stationed nearby and conducting
reconnaissance), positive arguments (suggestion of
a U.S.Vietnamese alliance against Russia and
China), or modification of position (a federal Vietnam
with South Vietnamese rulers maintained in power
through elections).

Suppose, as the UN commander, you want to eliminate
a trust dilemma without substantially changing your
position. You proceed to assess evidence and
arguments produced by the other side (the ethnic
militia) to prove it does not intend to defect. As you do
so, it will be inherently harder to eliminate your dilemma
(a) the more mistrustful you are and (b) the greater
your estimation of the cooperation dilemma the other
would face if it were to accept your position (or does
face, if it has already accepted it).

Chapter 3


These are distinct sources of friction. The first,

mistrustfulness, should be avoided. A highly
successful businessman said, I trust everyone until
they give me reason not to. He was explaining how
to make money out of a tendency to be over-mistrustful
(arising from the fact that in modern society we
frequently deal with strangers).
Nave, mistaken acceptance of others assurances is
best avoided, not by mistrustfulness, but by realistically
assessing the reasons for their cooperation dilemma
and helping to counter them in a positive spirit of
cooperation, even if, as noted, you are asking them to
cooperate with damaging acts of automatic retaliation
such as jets bombing roadblocks.
The bad effects of mistrustfulness are that, instead of
encouraging others to genuinely adhere to your
position, it invites them to treat it as a position that will
have to be abandoned. By projecting a skeptical,
detached attitude and treating your trust dilemma
(which is their putative cooperation dilemma) as
something hard to overcome, you give the impression
that you consider your own position untenable.

Willing Changes of Position

I actually may prefer to move, with others, to another
position. If this other position does not suffer from the
trust dilemma I face at my present position, I may
welcome my trust dilemma because it gives me an
argument for switching.
Suppose the wife actually wants to leave her husband,
the drunk. Her position, that she will stay if he will stop
drinking, is one she would be happy to relinquish. She


Confrontation Analysis

will then welcome her trust dilemma as giving her a

reason to leave him. Whatever he says or does she
will not want to trust him. Her mood will be skeptical
and uninvolved so as to deflate any emotion and
argument he may use to try to persuade her.
Suppose the local commander (belonging to a
coalition nation that lacks high standards of military
training) has been bribed to let the roadblock stay.
His position, that it be moved, is then one he wants
to give up. He accepts with detachment his trust
dilemma, the fact that he cannot trust the local militia
not to put back the roadblock as soon as his back is
turned. He deduces from it that there is no point trying
to get the roadblock moved.

Summary of Chapter 3
To win a confrontation by getting others to accept his
position, a commander needs to understand how
changes come about at the Climax phase. Gametheoretic dilemmas are the key. They face players that
try to pursue fixed objectives within a fixed frame,
causing them to feel emotion and to seek reasons to
change their preferences and the frame. The metaphor
of dramatic interaction seems better than gameplaying for describing such processes of
transformation; hence confrontation analysis is based
on drama theory, an extension of game theory.
Through trying to influence each other by arguments
in the common interest, dramatic characters tend to
build up the attitudes and interests of a supercharacter.
Six game-theoretic dilemmas pressure characters at
a moment of truth: the dilemmas of cooperation, trust,

Chapter 3


deterrence, inducement, threat, and positioning. The

cooperation dilemma faces a character that belongs
to a group that can do better by defecting from it (the
characters) own position. It is the dilemma of a drunk
trying to persuade his wife he means to stop drinking
after she has threatened to leave him if he does not.
Eliminating this dilemma without greatly changing
ones position means making an incredible promise
credible. It requires positive emotions and
rationalizations as to why you do not now prefer to
defect. Such protestations may be deceitful, inviting
disbelief, which finally can be overcome only by sound
reasoning and evidence. Friction to be overcome
involves hard-to-change values and beliefs. Some
friction is relatively slight, like that involved in keeping
easily-kept promises. Promises involving high stakes
need to be guaranteed with convincing reasons,
conversions, or sanctions.
One way of overcoming a dilemma is to change your
position to one that does not suffer from this dilemma
(though it may suffer from others). If you are committed
to your position, emotions of sorrow and despair may
help you to rationalize the switch to another position.
You may already prefer the other position, in which
case you may use the dilemmas of your present
position as arguments in favor of a move.
The trust dilemma faces a character when a group
not including the character is tempted to defect from
its (the characters) position. If it does not wish to give
up its position, it must eliminate the trust dilemma by
eliminating others temptation to defect, either by
arranging sanctions against it or by reasoned
persuasion and modification of its position to others


Confrontation Analysis

benefit. The required emotional tone is positive, even

when arranging punishing sanctions. Deceit is a
temptation and disbelief problem. Mistrustfulness is
counter-productive because it is negative in tone.

Chapter 4
More Pressure:
Dilemmas of Deterrence,
Inducement, Threat, and
The Deterrence Dilemma

face a deterrence dilemma in relation to a character

that is under no pressure to accept my position. It
feels no pressure because the threatened future does
not deter it from rejecting my position; rather, it
encourages it to.
In card-table terms, I have a deterrence dilemma in
relation to a character opposed to my position that
prefers the threatened future to it (my position).
Unlike the cooperation and trust dilemmas, this
dilemma arises only when parties take different
positions. There are conflicting positions in the local
commanders confrontation (see table 3), but not in
the problem faced by his superior (see table 2). In
table 4, the drunk and his wife share a position, but it
conflicts with that of the friend.
The high-level, general confrontation in the former
Yugoslavia between the Serbs and the West placed
the West in a deterrence dilemma for several years.



Confrontation Analysis

The West wanted the Serbs to desist from ethnic

cleansing. If the Serbs refused, the West threatened
to impose economic sanctions, and actually carried
out this threat. The economic sanctions were no
deterrent because the Serbs as a whole (taking into
consideration the dynamics of the interaction between
the Bosnian Serbs and the Serbian government)
preferred the threatened future, Ethnic cleansing with
economic sanctions, to No cleansing, no sanctions,
which was the Wests position.
Consider the local commanders confrontation in table
3. The local militia has not yet taken a definite fallback
position; therefore, what is represented is a point in
the Buildup phase of figure 2 at which both parties,
the local commander and local militia, have taken up
positive positions. The commander wants the
roadblock removed; the militia wants it to stay. Also,
the local commander has effectively stated the
following fallback position: If you dont remove the
roadblock, Ill forcibly remove it. If you then fire on my
troops, Ill call in air support. The local militia has not
yet stated a fallback position. Whether a deterrence
dilemma now follows depends on the fallback position
the local militia chooses. A refusal to remove the
roadblock (without a threat to fire on Allied troops) puts
the militia in a deterrence dilemma because the local
commander prefers the threatened future t (forcible
removal of roadblocks) to M, the militias position;
therefore, this would not pressure the commander to
accept M.

Chapter 4


Eliminating the Deterrence Dilemma

The deterrence dilemma, like other dilemmas, may
be eliminated by giving up ones position. For
example, the West might have agreed to let the Serbs
have their way, leading to most of Bosnia being
divided up between Serbs and Croats, with largescale genocide of Muslims. Such a withdrawal policy
was advocated by many in the West. The local militia
could solve its deterrence dilemma, if it had one, by
agreeing to remove the roadblock.
Suppose, however, that a character is determined to
retain much of its position. An alternative way of
eliminating the deterrence dilemma is to escalate to a
higher level of retaliation to make the threatened future
worse for the other party. This generally involves
thinking up, or bringing onto the agenda, new cards
with which to punish them. This process of thinking
up and making credible new punishment cards is
driven by negative emotions such as anger and
indignation. It is rationalized by demonizing the party
to be deterred; that is, seeing them as wicked and
evil, so that it becomes right and necessary to consider
extreme reprisals against them.
In the overall confrontation between the West and the
Serbs, Western public opinion demonized the Serbs
(while ignoring, or paying less than proportionate
attention to atrocities committed against them by
Croats and Muslims) until it was possible for serious
military intervention against them to become a credible
part of the threatened future, eliminating our
deterrence dilemma.


Confrontation Analysis

The West speedily demonized Saddam Hussein (for

example, relative to President Assad of Syria and
others) when he invaded Kuwait, and it became
necessary to get rid of a deterrence dilemma by
threatening military force against him.
In table 3, the militia can get rid of its deterrence
dilemma by adopting the threatened future tI, rather
than t; that is, by threatening to fire on Allied troops.
Though this invites retaliation from the air (this being
part of the commanders fallback position), it cures
their deterrence dilemma because the danger of allied
casualties means the commander prefers M (the
militias position) to tI. In making this threat, they are
likely to be driven by feelings of resentment and anger.

Escalation and De-escalationConciliation as a

Response to the Deterrence Dilemma
Thinking up new, more damaging threats is a
negative, escalatory way of resolving the deterrence
dilemma. It may be necessary, but it is risky. It runs
the risk of triggering a process of mutual
demonization and mutual escalation, in which each
party responds to threat escalation by further
escalation, driven at each stage by increasing
feelings of hatred and anger rationalized by viewing
the other as more and more evil.
This risk is greatest when characters have roughly
equal ability to escalate their threats. Clausewitz
(1968; 1 st edition 1832) generally assumes war
between equally matched states. Accordingly, he
sees mutual escalation as an inherent tendency of
war. By definition, he thinks war implies a sort of
reciprocal action, which must logically lead to an

Chapter 4


extreme in which even the most civilized nations

may burn with passionate hatred of each other (p.
103). Clausewitz, however, does not see why such
passions are aroused; that is, he does not see their
drama-theoretic function.
As we saw in chapter 1, war today is not generally
Clausewitzian. One side (the United States and its
allies) normally has superior capacity to escalate its
threats. This imbalance may diminish the risk of mutual
escalation, often enabling the allied side to cure its
deterrence dilemmas by one-sided escalation,
provided it is done in the right manner.
Consider a positive, conciliatory way of resolving the
dilemma. This is to add new cards that improve
(sweeten) our position in the estimation of the
threatened party. In this way the threatened future
becomes worse for them than our position, not
because we have made it worse, but because we
have made our position better. The emotion that goes
with this method is the positive one of reaching out
and sympathizing with the other s needs.
Concomitant rationalizations are designed to
accomplish the following:
Prove to ourselves that others have genuine
needs that deserve to be met
Establish our position, as now modified and
improved, as a win-win outcome that they ought
to accept.
Following are some examples:
The local militia, rather than escalate by
threatening to fire on the local commanders


Confrontation Analysis

troops, might try to persuade him to accept the

roadblock as a means of keeping peace in the
area. Well always let UN troops through, the
militia might say. We can use the roadblock as
a way of helping you accomplish your mission.
In February 1994, U.S. envoys reportedly offered
President Tudjman of Croatia a deal: stop
fighting the Muslims of Croatia and we will let
you take back the Krajina region from the Serbs
(Silber and Little, 1996). The ensuing de facto
alliance between Croats and Muslims led to a
succession of setbacks for the Serbs that
persuaded the Bosnian Serbs to accept the
Dayton agreement. Thus, the United States
shifted its position in favor of the Croats to
persuade them to stop fighting the Muslims, and
join them in fighting the Serbs.
The 1978 Camp David agreements between
Israel and Egypt brokered by U.S. President
Jimmy Carter were made possible by the United
States sweetening its position for both sides by
promising each of them large sums in aid.
By adding new cards that sweeten our position, we
do run the risk of creating a cooperation dilemma for
ourselves. Others may suspect we will not keep our
promise. If this occurs, we must use previously
described positive-emotion methods for overcoming
a cooperation dilemma by giving reasons why we
should be trusted. Our cooperation dilemma is their
putative trust dilemma.
Note that conciliation and trust-creation both involve
positive emotion. Following are the differences:

Chapter 4


Conciliation consists of modifying our position to

make it more attractive to others, thereby
eliminating a deterrence dilemma
Trust creation eliminates a cooperation
dilemma by giving reasons why we should
carry out our promise (i.e., why we should
prefer our own position to a possible
temptation to defect from it).
Conciliation needs to emphasize how our position
benefits them, trust creation how it benefits us.
Common-interest arguments have the virtue of
emphasizing both points at once.
Conciliation (i.e., enhancing our position from the
viewpoint of the other side) also has been discussed
as a method of getting rid of the trust dilemma. The
difference is that removal of the trust dilemma
addresses a situation in which both parties share the
same position, as with the UN theater commander in
negotiations with the ethnic militia. It assumes either
that this is the situation or that the trust dilemma is the
main objection to making it so (as with U.S.
negotiations with North Vietnam). Removal of the
deterrence dilemma, by contrast, tries to rectify a
situation in which the other side is refusing to accept
our position because they prefer the threatened future.
It is a mistake to confuse these situations.
U.S. policy in negotiating with North Vietnam, if our
analysis is correct, made just this mistake. It dealt with
a trust dilemma (i.e., How can we trust them not to
invade the South after we leave?) using methods
appropriate to a deterrence dilemma (bomb them until
they agree not to invade the south). These methods


Confrontation Analysis

brought the North Vietnamese to the conference table,

but left our trust dilemma unresolved.

Conciliation Combined with Escalation

Conciliation as a method for overcoming the
deterrence dilemma has a disadvantage; it may mean
we must make distasteful concessions. Conciliation
became particularly discredited by British Prime
Minister Chamberlains attempts to conciliate Hitler in
the Munich crisis of 1938; conciliation was famously
used then as an alternative to escalation, because
Britain and France were unwilling to threaten war.
Conciliation may be effectively combined with
escalation in a manner that makes large concessions
unnecessary. Pure escalation has a negative
emotional tone and rationalization; it tends to
rationalize a preference for anything that is against
the others interests. Consequently others will fear that
their interests will not be safeguarded in any settlement
they might discuss with us. (Recall that the exact
details of any settlement are discussed in the
Resolution phase following overall agreement.)
Consequently, rather than discuss a settlement at this
stage, they will look for possibilities of counterretaliation to lay the basis for a more balanced solution.
In this way, pure escalation encourages counterescalation, if possible.
Combining escalation with conciliation in a tough cop,
tender cop routine, gives the impression that while
determined to punish refusal to settle, we are willing
to be sympathetic to others interests if they do settle.
This impression may count for more than any concrete
concessions you offer, allowing actual concessions

Chapter 4


to be relatively slight. The point is that others need

negative feelings and rationalizations directed against
us to take the path of escalation. Our concern for their
interests undermines such negative feelings and
makes it harder for them to justify escalation. In terms
of the internal confrontations that determine their
policies, conciliation gives the doves within them
arguments against the escalation-favoring hawks.

The deterrence-dilemma friction consists of any
inherent difficulty in making others believe that the
threatened future is worse for them than your position.
This friction may be great or small.
For example, in 1995 at Wright-Patterson Air Force
base in Dayton, OH, Holbrooke took Milosevic and
Bulatovic (the Montenegrin President) into the
Nintendo room, a map center equipped with
computers that allowed the user to overfly the Bosnian
terrain. At this point the positions of the Serbs and
the United States were close, differing only in regard
to whether a certain area should be under Serb
control. Showing him the area in dispute, Holbrooke
said, Am I seeing right? Theres nothing there. Just
mountains. No houses, no villages. Bulatovic said,
Thats right, but this is Bosnia. Holbrooke
responded, Look at what youre fighting for. There is
nothing there. He was producing reason and
evidence in favor of the U.S. position by showing the
Bosnians how close it was to their own, and hence
how much they should prefer it to the threatened
future. (See Silber and Little, 1996, p. 373.)


Confrontation Analysis

When at the Athens meeting in 1993 Cyrus Vance

told the Bosnian Serbs the U.S. Air Force is all
prepared to turn Bosnia and Serbia into a
wasteland, he was giving reasons why they should
consider the threatened future to be worse for them
than the U.S. position.
Overcoming this type of friction is a matter of working
on others beliefs and values to affect their comparison
between our position and what they can hope for if
they reject it. Making our position or fallback position
credible, given the extra cards we may have added to
them, is a separate matter of solving either a
cooperation dilemma or one of the three remaining
dilemmas examined below. It is not a matter of the
deterrence dilemma as such.

Subcharacters Aiming to Change Your Position

We have said you can eliminate a deterrence dilemma
in a conciliatory manner by changing your position in
the sense of giving up something; however, such a
change of position may be just what certain
subcharacters belonging to your organization prefer.
They may use your deterrence dilemma as an
argument in favor of the position change that is their
For opinion holders in the West who preferred
not to oppose the ambitions of Hitler, the Soviet
Union, China, North Vietnam, the Serbs, and
Saddam Hussein, the Wests deterrence
dilemma (How can we deter them?) was not a
problem but an opportunity. They used the
answer, We cant, as a reason to stop

Chapter 4


Similarly, for opinion holders in the 1980s in the

Soviet Union who wanted the Soviets to move
toward the Wests position, the Soviet
deterrence dilemma (How can we match the
Reagan arms buildup to negotiate from
strength?) was not a problem. They too would
have argued, We cant. Thats why we should
give in.
When internal confrontations take place to determine
the policies of large characters such as the West or
the Soviet Union, a subcharacters true objectives may
be revealed by its attitude toward dilemmas. One that
really wants to achieve the organizations stated
position will show negative emotion in face of a
deterrence dilemma. One whose true aim is to change
that position will be more detached and objective.
Detachment and objectivity go with a tendency to
regard the frame as fixed, encouraging positional
change because if the frame is fixed, then the only
way to escape from a dilemma is to change position.

The Inducement Dilemma

The inducement dilemma is the other side of the
deterrence dilemma: In eliminating my deterrence
dilemma, I give you an inducement dilemma.
In card-table terms, a character has an inducement
dilemma if anothers position (though different from
its own) is as good for it as the threatened future.
It follows that by successfully solving a deterrence
dilemma I give someone else an inducement
dilemma. However, I may give myself one as well.


Confrontation Analysis

When the West, in its overall confrontation with the

Serbs, finally started threatening armed intervention,
it replaced its deterrence dilemma with an inducement
dilemma. It preferred the Serbs position (We continue
ethnic cleansing) to implementing its new fallback
position (i.e., attacking the Serbs). This became
obvious to the Serbs as the West, in one
subconfrontation after another, failed to follow up on
its threats. In terms of the five possible endings in figure
2, each confrontation ended with the implementation
of a threatened future flunked by the West.
After Munich, as Hitlers ambitions proved to be
unappeased, Britain and France started to rearm,
thereby making war part of their threatened fallback
position. In this way they may have eliminated their
deterrence dilemma (Hitler might have preferred to
back off rather than fight) but replaced it with an
inducement dilemma (Hitlers disbelief in their
willingness to fight).
These examples illustrate that while my deterrence
and inducement dilemmas may be similar in their
effects (in that both may allow my opponents to have
their way undeterred by me), they are different in their
causes and need to be tackled differently. My
deterrence dilemma is a matter of my opponents
preferences. I must ensure that they prefer to accept
my position rather than provoke me into taking up my
fallback position. Having ensured this, I may still face
an inducement dilemma, which is a matter of my own
preferences. The question here is, do I prefer to give
in to them rather than implement the threatened future?
If so, I have an inducement dilemma to overcome to
make my deterrence fully credible.

Chapter 4


Avoiding Escalation by Accepting the

Inducement Dilemma
By overcoming my inducement dilemma, I give the
other side a deterrence dilemma. This forces them to
choose between giving in or overcoming that dilemma.
The latter choice gives me another inducement
dilemma to overcome, and so on. For both sides to try
to overcome both dilemmas means embarking on a
cycle of escalation.
Therefore the question arises: Is it possible to
negotiate in a situation when both sides have an
inducement dilemma and neither has a deterrence
dilemma? The answer is yes.
Under the nuclear deterrence regime of the Cold War,
both sides were in this position. Each must have
preferred, at each confrontation, to give in to the other
rather than start a nuclear war. Because this situation
was clearly symmetrical (balance of terror), neither
side simply gave in; they negotiated.
Hitler and Stalin negotiated the NaziSoviet pact in a
similar way. Each preferred a range of positions to a
RussoGerman war because for each of them those
positions meant a license to gobble up their neighbors,
their military forces being the only ones in Europe that
presented a serious threat. (Each knew, of course,
that after this gobbling-up process had reached its
limits, all bets would be off.)
Often in such negotiations the threatened future is
tactfully left unmentioned by the parties. They
negotiate positively, advancing their own positions and
attacking others on the grounds of each positions
contribution to the common interest. Because all


Confrontation Analysis

positions being discussed are preferred to the

threatened future by all parties, continually comparing
them with that threatened future is unnecessary.
Balanced negotiations of this kind, in which each side,
rather than suffer a breakdown, would prefer to accept
a range of positions, are the norm in civilized relations
(e.g., most trade negotiations). Such negotiation
situations are not stable. They contain dilemmas,
causing them to move; however, they are prevented
from escalating and kept within stable bounds by
directing the destabilizing effect of the inducement
dilemma toward impelling characters to negotiate a
single position. This is a way of eliminating the
deterrence dilemma that is a non-escalatory alternative
to preferring the threatened future to others positions.
After they have agreed to a single position, the
characters generally will face cooperation and trust
dilemmas. Foreseeing such dilemmas will be one of
the factors leading them to accept this or that position.
Ideally the main factor leading them to agree will be
passionately argued rational arguments in the common
interest, even when the common interests involved
are as crude as in the case of the NaziSoviet pact.

Other Ways of Eliminating the Inducement

Should we then see balanced negotiations (i.e., initial
acceptance of the inducement dilemma followed by
rational, common-interest debate) leading to its
elimination through convergence to a single position,
as the norm for defense forces. Should we aim for it?

Chapter 4


We probably should when considering relationships

between internal actors (the subcharacters within an
alliance whose confrontations determine its policies).
We are then concerned with cooperative
confrontations, as discussed in chapter 10. We
probably should not when our defense forces face
rebels against the New World Order. In such cases
we typically find an extremist approach on the rebel
side (i.e., a willingness to escalate threats to a point
where they can dictate the outcome), and an
asymmetry of potential power with the Allies, led by
the United States, having superior escalatory capacity.
Opponents motivated by a philosophy requiring them
to refuse any kind of compromise may reject the
idea of negotiating on the basis of common interests.
They may work themselves up into preferring a
breakdown (i.e., the threatened future), no matter
how dangerous, over any position acceptable to us.
Our forces then may need to eliminate our
deterrence dilemma by putting on the table cards
that both punish and conciliate the other sufficiently
to make them prefer at least one acceptable position
to a breakdown. This means combining adequate
deterrence with a conciliatory posture. It means
carefully lining up our guns to point at their head,
then saying, Right, lets talk about your problems.
It also means eliminating our inducement dilemma
by preferring the breakdown (firing our guns) to
acceptance of their extreme position.
Eliminating our inducement dilemma gives them a
deterrence dilemma, forcing them to either modify their
extreme position or escalate by making the breakdown
still worse for us. We are trying to ensure the former


Confrontation Analysis

response. To forestall the latter, we must eliminate

our inducement dilemma so thoroughly as to prefer
the breakdown to their extreme position, no matter
how much they escalate. In general we have the
capacity to do this because of our preponderance of
military power, although it is true that new threats of
chemical or biological attack on cities may enable
terrorist groups to make the West back down.
U.S. and Allied preparations for the Gulf War, both
military and psychologicalpolitical, exemplified
thorough elimination of our inducement dilemma; it
covered all possibilities of Iraqi retaliation and prepared
us for the worst. The problem seems to be that we
failed to cure our deterrence dilemma (i.e., Saddam
Hussein still preferred war to leaving Kuwait). We may
have assumed too much common interest between
him and the Iraqi people, who were the ones that
suffered from the war.
What, then, is the recipe for eliminating an inducement
dilemma when the other side will not budge from an
unacceptable position? Negative emotion and
rationalizations are needed; however, they do not need
to be directed entirely at the other side as they must
be to eliminate the deterrence dilemma, where we
need to think of cards that will hurt them. It is now our
own preferences and underlying values that concern
us. We need to think sufficiently poorly of the extremist
position being rejected and reconcile ourselves
sufficiently to the threatened future, to prefer the latter
to the former. It can be a case of hate the sin (the
position being rejected), not the sinner (the characters
taking that position). A feeling of martyrdom (By
taking this position they are forcing us to accept the

Chapter 4


threatened future) may be the greatest negative

feeling invoked against the other side.
The major argument leading a wide coalition of nations
to support Operation Desert Storm was the
unacceptability of letting Saddam Hussein get away with
the annexation of Kuwait. Many reasons were given for
this, taking account of different parties interests.
Hostility to Saddam Hussein was not important, except
for the Allies who had to gird themselves to put the
intervention card on the table. They were the ones who
had needed to make an effort (unsuccessful though it
turned out) to overcome the deterrence dilemma as well
as the inducement dilemma.
Modifications or reappreciations of the threatened future
that make it more acceptable to us, but not to the other
side, also help eliminate our inducement dilemma.
For example, at the outbreak of World War I, as to
some degree with any hostilities, there was patriotic
pride and excitement at the prospect of using violence.
This precisely fitted the bill, making the threatened
future more attractive to us while encouraging creative
ideas as to how to make it worse for them.
It might seem there also could be modifications of the
other sides position, as distinct from reappraisals of
it, to make it less attractive to us. There are suspected
cases of this. For example, the suspicion that President
Franklin Roosevelt permitted Pearl Harbor to happen
or that in February 1994 Muslims planted the bomb
that blew up their own people in Sarajevo. Here, covert
direct intervention is suspected of having made a
position worse in someones eyes, thus overcoming
an actual or potential inducement dilemma. Why must


Confrontation Analysis

it be covert? It cannot be done openly because it is

inconsistent to encourage something that we
denounce (i.e., evil aspects of the others position).
Inconsistency is against reason, and reason is
necessary to compel belief at a moment of truth. We
admire inconsistency and confessed self-manipulation
in our leisure moments, not at the moment of crisis in
a confrontation.

Friction and Subcharacter Conflictsthe

Inappropriateness of Cost-Benefit Analysis
The inducement dilemma may be hard to overcome,
even for a character that can easily overcome the
deterrence dilemma. It is relatively easy for the United
States, leading the West, to threaten action against
tyrants and murderers; it is harder to carry it out,
particularly when U.S. citizens lives may be lost. This
can lead to a commander being given absurd
mandates such as, Threaten to do it; but dont do it
without permission.
What is needed is the political will to take action if
necessary, supported by an understanding that such
willingness to carry out threats will mean that they will
less often have to be carried out (provided the
deterrence dilemma is first overcome, so that the
threats are adequate). There may always be
subcharacters within our character that want to shift
position toward acceptance of an opposing position
that is giving us an inducement dilemma. They will
point to disadvantages of the threatened future as if
our preference for it should be determined by costbenefit analysis. But this tool is inappropriate when
comparing the threatened future with a position we

Chapter 4


are rejecting, because we then have a paradoxical

need to prefer to carry out our threats in order not to
have to carry them out. Concepts such as honor,
patriotism, solidarity, ethics, and adherence to principle
meet this need. These emotional concepts have no
role in economics, but they do have a vital role in
resolving confrontations in which military action is a
card that needs to be made credible.

The Threat Dilemma

Often the threat dilemma coincides with the
inducement dilemma, but it is conceptually and
practically different. It occurs when I cannot be trusted;
I cannot even trust myself to implement my part of the
threatened future if we move into the conflict phase of
table 2.
In card-table terms, I face a threat dilemma when I
can move, just by changing my selection of cards, from
the threatened future to another future I like just as
well. Therefore, at the moment of truth, others will
suspect that I am bluffing.
We saw in chapter 2 how the West flunked the
implementation of Vances threat to turn Bosnia and
Serbia into a wasteland made in May 1993 to the
Bosnian Serbs in Athens. The West faced a threat
dilemma: it preferred doing nothing to launching a
bombing campaign. It is unclear if the Serbs suspected
them of bluffing, but the failure must have damaged
Western credibility for the future. (See Silber and Little,
1996, p. 282.)
The dilemma often coincides with the inducement
dilemma because the temptation I feel to refrain from


Confrontation Analysis

implementing the threatened future may be a

temptation to accept your position. However, I may
have other temptations to defect to from the threatened
future. In any case, the two dilemmas are different.
The threat dilemma is like the cooperation and trust
dilemmas in that it arises from parties looking forward
to what is likely to happen in the Implementation
phase. It is a problem for me because others suspect
I might not prefer to carry out my threat if and when
the time comes to do so; therefore, they discount my
threat as incredible.
The inducement dilemma, like the deterrence dilemma,
is grounded in the tug-of-war that takes place at a
climactic moment of truth. I am under pressure then
and there to give in to your position, because to do so
is as good for me or better than the threatened future;
therefore, for purposes of the inducement-dilemma
argument, the threatened future is assumed to be
credible, not incredible. By contrast, my threat
dilemma is all about questioning the credibility of the
threatened future. You are disinclined to believe I will
carry out my threat. If I walk out threatening to do so,
you think I will either come back, having changed my
mind, or do something else instead.

Getting Rid of the Dilemma

Despite this difference between the two dilemmas,
many of the emotions and rationalizations we use to
overcome an inducement dilemma work as well for a
threat dilemma, for the simple reason that both can
be overcome by raising our valuation of the threatened
future. In this way we overcome the inducement
dilemma by raising its valuation relative to anothers

Chapter 4


position; we overcome the threat dilemma by raising

it relative to any temptation we have to defect.
Helpful emotions are defiance, anger, indignation, and
the martyred feeling of being forced into the threatened
future by others intransigence. Temptations may be
downplayed, with reasoning as to why we would not
want to do that anyway. The most general, all-purpose
rationalizations center on elevating our preference for
the threatened future. There are two kinds: first,
evocation of principles such as honor, self-respect,
integrity, and the need to keep commitments; second,
evaluation of our fallback position as being
instrumentally the best way, or at least a good way, to
deal with the situation created by others resorting to
their fallback positions.
The British continually have used both kinds of
argument to justify their continuing fight against
terrorism in Northern Ireland, despite the publics
frequently-stated preference for leaving them to fight
it out. The same two kinds of argument were used in
the United States to justify the continuing involvement
in Vietnam and, more recently, the decision to fight
Saddam Hussein.

Subcharacter Conflicts Over the Threat Dilemma

It is essential to note that friction to be overcome in
eliminating the threat dilemma is felt in the
Implementation phase itself. Its effect at the moment
of truth or in the preparatory Conflict phase is
derivative. It derives from anticipation of the problems
that will or would arise during implementation.


Confrontation Analysis

In the Implementation phase we are faced with

needing to carry out our threats, as the United States
did in launching Operation Desert Storm, or failing to
do so, as when the United States could not convince
the Europeans they should follow through on Cyrus
Vances threat to turn Bosnia and Serbia into a
wasteland. Hard questions of ethics and self-interest
arise at this point that were glossed over when making
the threat. Many who previously supported our position
now change their minds and join those who were
always against it in arguing that we should not carry
out our threat. As in the case of the inducement
dilemma, such opposing subcharacters may use costbenefit arguments. Threats having failed, it may seem
that the realistic thing to do is to abandon our position.
The point is that these difficulties are foreseen,
accurately or inaccurately, and possibly with
exaggerations in either direction, by those who, at the
moment of truth, are faced with assessing the
credibility of our fallback position. The more difficulties
they foresee for us, and the more reasonable they
think we will be in succumbing to them, the less
credible our threat, and the more likely it is that we will
be faced with the hard choice of put up or shut up.
The Argentineans were amazed when the British fleet
set out to recapture the Falklands, having confidently
foreseen that Britains threat to do so was futile.

The Positioning Dilemma

The positioning dilemma occurs when I prefer
anothers position to my own. In card-table terms, I
like the others position column more than mine.

Chapter 4


This is likely to occur when, under pressure of

dilemmas, I have abandoned a position I occupied
with others and accepted one that I find less preferred,
leaving others with whom I shared the first position to
continue to occupy it.
A player might abandon a preferred position in favor
of one it likes less because of arguments based on
realism. At its old position, the player faced dilemmas,
which, unlike the players that still occupy that position,
it came to consider insurmountable; however, now it
faces a dilemma in arguing with remaining proponents
of its old position because it prefers what it is arguing
against to what it is arguing for. The player stands
accused of insincerity and dishonorable conduct in
failing to stand and fight for what it once believed in.
The accusation is one of failure to live up to principles
the player previously defended. A common example
is that of two friends. In their youth they supported
left-wing causes. In middle-age, one continues to
support left-wing causes while the other argues that
the positions they once shared are unrealistic.
Feelings of guilt at having failed to live up to oncedefended principles accompany this dilemma. These
guilt feelings arise out of the dilemma. They are
structural. We are wrong to think of guilt feelings as
being caused by the abandonment of higher feelings
such as altruism in favor of lower ones such as
selfishness. The same guilt is felt in response to a
positioning dilemma in which lower principles are
abandoned in favor of higher ones. To see this,
consider two middle-aged friends who thieved and took
drugs in their youth. One has reformed, primarily on
realistic grounds (deterred by societys punishments


Confrontation Analysis

for such behavior), but also on the grounds that the

behavior is wrong. His friend will accuse him of
deserting lower principles for higher ones, and if his
main reason for changing position is realism, he will
feel guilty in face of this accusation.
The positioning dilemma is essentially a guilt dilemma.
It is overcome by a change of principles (or a
reassignment of weights between different principles)
that eliminates the dilemma by making the new position
preferred to the old one. The emotion accompanying
this change is the zeal and enthusiasm of a convert,
rationalizing negative attitudes toward the old position
and positive ones toward the new. Former
Communists become zealous defenders of freemarket capitalism. Former drug addicts campaign
vehemently against drugs.
Often such converts are respected for their supposed
inside knowledge of the position they are attacking.
Actually, their knowledge may be biased, having been
through strong emotional pressures.

Modeling Northern Ireland Negotiations

To see how the positioning dilemma may affect peace
operations, we will model the negotiations in Northern
Ireland preceding the talks that led to the Easter 1998
agreement. Here we will see a number of dilemmas
at work, in addition to the positioning dilemma.
Table 5 shows the moment of truth between Britain,
the Irish Republican Army (IRA), and the Unionists
before the reconvening of peace talks in 1997. It
focuses on the point when a new government under
Prime Minister Tony Blair took over from the

Chapter 4


Table 5. Status of Northern Ireland

peace talks when Tony Blair took over.

Conservatives, under which an IRA cease-fire broke

down in early 1996. This cease-fire was meant to
precede all-party peace talks. It broke down over the
British governments insistence on backing the
Unionist demand that Sinn Fein (the IRAs political
wing) not be admitted to talks unless the IRA first
handed in some of its weapons.


Confrontation Analysis

When Blair took power, the parties positions were

as shown in table 5. In column I, the IRA was offering
to cease fire (but not disarm) if it could then be
admitted to peace talks. In column U,B, the Unionist
position, shared by the British, was that Sinn Fein be
admitted to talks only if they disarmed. If this was
rejected, the British and Unionists were threatening
to go ahead with talks without the IRA (in fact, talks
were ongoing, but marking time). Thus the threatened
future consisted of the IRA neither ceasing fire nor
disarming, the British not admitting them to talks, and
the Unionists not quitting talks. The default future was
the same.
In this table, as before, the numbers beside characters
names represent their preference rankings for the
various futures, derived from the values they are trying
to pursue in this confrontation. Number 1 is attached
to the most preferred future, 2 to the next most
preferred, and so on. From these preference rankings,
some of the dilemmas can be read off. Others are
indicated by question marks (i.e., a card has a question
mark on it if there is some doubt if it would actually be
played or not played).
For example, the distrust felt by each side toward the
other is indicated by question marks beside the
Unionists Not quit talks choice in column I and the
IRAs Cease fire choice in column U,B. This is
because the IRA suspected that if talks took place on
fair terms, as in column I, the Unionists would sooner
or later quit such talks, while the Unionists suspected
that the IRA would sooner or later break the cease
fire even if, as in column U,B, they previously had given
up some of their arms, which the Unionists in any case

Chapter 4


suspected they would not. (Even if they gave up most

of them, the Unionists suspected, the IRA could always
rearm.) This means that both the IRA and Unionists
have trust dilemmas. Each also has a potential
cooperation dilemma because each can envisage talks
going in directions so unacceptable that they would
prefer to return to violence or quit the talks.
These dilemmas gave the characters arguments to
reject each others positions. At the same time, each
was inclined to regard its own position as second-best
and to argue that, in light of the above dilemmas, the
British should move with them to another position, one
in which the other side was excluded from negotiations.
For the Unionists, this preferred position was the
default future d, the same as the threatened future t.
For the IRA, it was the same as their position I, but
with the Unionists provoked into playing their Quit
talks card.

Britains Dilemmas
What caused movement was the combination of
dilemmas faced by the British. They had not only an
inducement dilemma (they preferred I to t), but also a
deterrence dilemma (the IRA, because they would find
it impossible to retain internal discipline if they
disarmed ahead of talks, preferred t to U,B).
It was the British deterrence dilemma that was
decisive. Unless you can get rid of this dilemma, you
are not in the game. Meanwhile, the IRA had neither
an inducement nor a deterrence dilemma. It was their
otherwise unsolvable deterrence dilemma that
precipitated the Blair governments change of position.
The Blair government accepted the IRAs position (i.e.,


Confrontation Analysis

that it should be admitted to talks without disarming).

In table 5, the government shifted position from
columns U,B to I.
The same deterrence dilemma faced the Unionists. It
had the effect on them too of making them want to
change position, but in a different direction. The
Unionists used their deterrence dilemma, like their
other dilemmas, to argue that they and the British
should take the status quo (the default future d, the
same as the threatened future t) as their position.
Why did the same imbalance not precipitate a change
of position on the part of Blairs Conservative
predecessors? The simple explanation, often given,
is that they depended on Unionist politicians to keep
their majority in Parliament, and hence stay in power.
This might not have been enough to keep them aligned
with the Unionists if a positioning dilemma did not lie
ahead if they changed position. Consider the situation
of these Conservative politicians. They preferred, for
good reasons, the position U,B to the IRA position I.
This meant they need not think about their majority in
Parliament. Instead they could think of all kinds of
principled reasons, such as not giving in to terrorists,
for staying with the Unionist position. Blairs
government, by contrast, went against its own
principles in shifting position.
The Blair government went against its principles but
embraced political realism. The solution U,B in table 5
was simply unrealistic. The Blair government might have
tried to make it realistic by escalating other threats to
the IRA, such as potential loss of U.S. and Irish support,
followed by a heavier crackdown on terrorism. The
Unionists suggested these extra threats, but the

Chapter 4


Unionists wanted the threatened future, and all such

threats were, in fact, things the Unionists wanted to
happen for their own sake.

New Moment of TruthBritains Positioning

When the Blair government shifted position, a new
moment of truth emerged, as shown in table 6. Britain
now shared the IRAs position. The IRA responded
by calling a new cease fire, without disarming, giving
a clear signal of support for what was now a joint
BritishIRA position. (Notice how this cease-fire, being
reversible, acted merely as a signal of intent, not as
an element of the implementation. Evidently it
remained possible that the implementation, when it
came, would not contain the cease-fire card.)
The threatened future was now that the talks would
go ahead without the Unionists. The Unionists always
said that if Sinn Fein was admitted to talks without
having disarmed, they would quit. Blair now called the
Unionists bluff, declaring that the train, which under
the Conservatives had threatened to leave without
Sinn Fein, would now leave, if necessary, without the
Unionists; however, in light of the Unionists declared
fallback position, the default future was now one in
which the Unionists would refuse to join the train (i.e.,
it was the same as the new threatened future).
This new threatened future naturally changed the
dilemmas parties faced. It was no longer the Unionists
that preferred the threatened future to their own
position; it was the IRA. The Unionists now faced an
inducement dilemma, not wanting to accept their
opponents position I,B. Yet the Unionists preferred


Confrontation Analysis

Table 6. Status of Northern Ireland talks after Blair shifted position.

this position to t, because the latter would have left

important decisions to be made in consultation
between the British government and their sworn
enemies. If the train was going to leave, they had to
be on it.
In addition, they now faced a deterrence dilemma.
Their position, U, which they now preferred to t,
remained unrealistic. It was still the case (more so, in
fact, than before) that t brought no pressure on the
IRA to concede. The IRA preferred t to U even more
than before. This deterrence dilemma gave the British

Chapter 4


an argument for refusing to accept U, even though in

doing so the British had a positioning dilemma. (The
government actually preferred U to its new position
To overcome their now-pressing deterrence dilemma,
Unionist militants argued in favor of escalation (i.e.,
campaigns of disobedience and disruption). Unionist
leaders, however, saw little future in becoming
lawbreakers and allowing the IRA to claim the mantle
of respectability.
Other dilemmas as before faced the IRA and the
Unionists, if their positions were to be accepted. The
IRA and the British now had no deterrence dilemma;
they were able to put pressure on the Unionists to
give in. Nor did the IRA have an inducement dilemma.
The British did have an inducement dilemma, and
worse, a positioning dilemma. In talking to the
Unionists, the British had to argue against a position
they preferred in favor of one they disliked. Their
argument could not, therefore, be based on any
supposed virtues of I,B as compared to U. They had
to admit that U was better. Their point was merely
that U was unrealistic.
The Unionists, meanwhile, could not rely on Britains
inducement dilemma. They might hope that the new
threatened future t in table 6 would prove unbearable
to the British, who would eventually be tempted to
break off talks with the IRA and resume them with the
Unionists. This hope was outweighed by the fear that
the Blair government would face no such dilemma.
The anguished Unionists finally accepted the position
I,B and all-party talks began.


Confrontation Analysis

How to Get Rid of the Positioning Dilemma

In this example, Britain eliminated its positioning
dilemma by getting the Unionists, whose position they
preferred to their own, to accept their position. The
government did so using the argument that the
Unionist position was unrealistic.
Undoubtedly, they must also have argued that their
position, which they shared with the IRA, was not
as bad as they had formerly painted it. As long as
the Unionists continued to reject realistic
arguments, British negotiators were under pressure
to eliminate their positioning dilemma, not just by
appeals to realism, but also by adjustments to their
beliefs and values.
The emotion that drives such elimination is
embarrassment at being unable to give meritorious
arguments for something that you are prepared to
accept for reasons other than its merits. This produces
negative emotion directed at the character you are
arguing with, together with rationalizations building up
the merits of the advocated position as against the
one rejected. In extreme form, this emerges in the
emotions of a convert, who generally feels more
strongly in favor of its new position and against its old
one than those who never took the old position.

Subcharacter Conflicts Over the Positioning

The friction involved in overcoming the positioning
dilemma is simply the difficulty you find in overturning
the beliefs and values by which you once justified your
old position against your new one. Scientists accepting

Chapter 4


a new theory, for example, must go methodically

through each of the old arguments to see why points
they previously rejected are now acceptable. Similarly,
converts to a new morality have to critically review old
life situations in which they made wrong choices.
This is what you must do to justify your new position.
Meanwhile, for subcharacters who think we should
have kept to the old position, our positioning dilemma
is welcomed as an argument against the new position.
It may similarly be welcomed by those who favor a
third position, as they can use the dilemma to point
out that we are being illogical.
When negotiations break down and characters enter
the Conflict phase of figure 2, each character generally
splits into subgroups, within which confrontations take
place between those who want to proceed with the
fallback position and those who want to restart
negotiations or otherwise flunk the Conflict phase. In
such a confrontation, those whose position is
Implement our fallback position may be accused of
inconsistency on the grounds that until now they have
advocated another solution, the one they were urging
at the Climax phase.
For example, following the breakdown of the Vance
Owen initiative in 1993, Secretary of State Warren
Christopher urged the European allies of the United
States to support lift and strike, (i.e., lift the arms
embargo and supply the Muslims with arms while
conducting air strikes against the Bosnian Serbs). Lift
and strike was the threat Vance and Owen had used
unsuccessfully to try to get the Serbs to agree to
VanceOwen. The European response was, We still
support VanceOwen. Why dont you?


Confrontation Analysis

Wanting to continue to press for the VanceOwen

position, the Europeans accused Vance of
inconsistency in having abandoned it. Owen himself
argued against the air strikes he previously had used
as a threat, saying, You will not solve the problem at
10,000 feet. The Europeans used what had now
become the Americans positioning dilemma, their
advocacy of lift and strike even though they preferred
VanceOwen, as an argument against the U.S.
position. The same argument was used by others who,
judging that VanceOwen could not be revived,
advocated withdrawal or other positions.
This example illustrates another point. What was the
threatened future during the Buildup and Climax
phases becomes a position to be taken up in the
Conflict phase. The future that you were using as a
threat, all the time hoping you would not have to carry
it out, becomes a position that you may have to argue
for against opposition from other subcharacters.

Summary of Chapter 4
I face a deterrence dilemma in relation to a character
opposed to my position, who prefers the threatened
future to my position. The problem I face is that my
fallback position does not pressure this character to
accept my position. I can eliminate my dilemma by
giving up my position. Alternatively, I can escalate to
a higher level of retaliation by thinking up new cards
to punish the opposing character. Anger and
indignation, leading to demonization of the opposing
character, accompanies this creation of new cards.

Chapter 4


Conciliation is another way of eliminating the dilemma.

To pursue conciliation, I retain the important
characteristics of my position while sweetening it until
the opposing character prefers it to anything it can obtain
from the threatened future. A tough-cop, tender-cop
routine is a way of combining escalation with conciliation.
Certain subcharacters within our character may
prefer us to abandon our position and accept that of
another player. They will tend to use a deterrence
dilemma in relation to that other player as an
argument for abandoning our position, on the
grounds that it is unrealistic.
I have an inducement dilemma if anothers position
(different from my own) is as good for me as the
threatened future. This puts me under potential
pressure to accept that position, rather than go to the
threatened future. I can overcome this dilemma by
finding reasons to prefer the threatened future to their
position, but in so doing I give them a deterrence
dilemma. I escalate the conflict between us.
To avoid escalation, each side must accept its
inducement dilemma. It must try to eliminate it by
working out a common position with the other side.
This means using passionately presented rational
arguments in the common interest.
Against an extremist opponent, the best strategy may
be, first, to eliminate our deterrence dilemma by mixing
conciliation with a threat sufficiently strong that they
cannot hope to escalate their way out of it; second, to
eliminate our inducement dilemma by preferring to
carry out this threat rather than accept any extremist
position; third, to negotiate an acceptable common


Confrontation Analysis

position that will not now be extremist. Under such

circumstances, cost-benefit analysis will be an
inappropriate tool for evaluating the threatened future.
Concepts such as honor, commitment, and adherence
to principle are appropriate.
I face a threat dilemma if I would be tempted to defect
from the threatened future, if it were implemented.
Unlike the inducement dilemma (with which it may,
but need not, coincide) the deterrence dilemma
affects characters at the Climax phase because they
look forward to how it would affect them if they moved
to the Conflict phase and considered whether to
implement a future (the threatened future) from
which they are tempted to defect.
I face a positioning dilemma when I prefer anothers
position to my own. This is likely to occur when I have
abandoned a position I occupied with others and
moved to one I find less preferred, but more realistic.
My dilemma will cause me to feel guilty when arguing
for the new position against the old, unless I can
change my principles sufficiently to prefer the new one.
Having made this change, I will, as a consequence,
tend to uphold the principles supporting my new
position even more strongly than those who have
always held it.
An alternative way for me to get rid of my positioning
dilemma is to get those who still hold my old position
to move with me to my new one. This is how the British
government eliminated the positioning dilemma it faced
when arguing with the Unionists in favor of admitting
Sinn Fein to talks although the IRA had not yet
decommissioned arms.

Chapter 4


A positioning dilemma may face subcharacters of a

character that has to put up or shut up (i.e., decide
whether to implement its fallback position after
negotiations break down). Subcharacters who
previously argued for internal implementation of a
peaceful solution now may find themselves arguing
for implementation of a punishing policy that was used
in the higher-level drama as a threat to pressure others
into accepting the peaceful solution. Other
subcharacters may accuse them of abandoning their
principles in switching from a peaceful policy of the
internal game to a punishing one.

Chapter 5
When All Dilemmas
Are Eliminated

hat happens when all six dilemmas are


We have seen that while any dilemmas remain,

characters are under rational and emotional
pressure to get rid of them; therefore, the situation
is not fully stable (i.e., there is discontentment and
pressure to change). This must continue until no
dilemmas are left.
Is the situation then fully stable, with fully contented
characters? This question is important for the
commander of a peace operation that has at least a
partial goal to bring about stability. The answer is yes,
with several qualifications.

The Final State

It is true that the final state, where there are no
dilemmas that give rise to pressures for change, is
stable; all characters agree on a single position, and
no character or group has any temptation to defect
from this position. This theorem of the final state is
proven in the mathematical appendix. It can be made
intuitively clear as follows:



Confrontation Analysis

If all parties agree on a single position, then the

absence of any cooperation dilemma ensures
that no character or group has any temptation to
defect from it.
If there is not a single position, then the only way
for characters to eliminate their inducement
dilemmas without creating deterrence dilemmas
for each other is to converge to a single position.
If there is a single position and no cooperation
dilemmas, then no other dilemma can exist
except the threat dilemma.
Note that the threat dilemma can still exist; however,
this is unimportant because if all agree on a position
that they can trust each other to carry out, then any
lack of credibility in what they threaten to do, if they
do not trust each other, ceases to matter.
The final state is, then, a completely satisfactory solution,
satisfying both emotion and rationality. Correct?
That is correct; however, we have said there are
qualifications to bear in mind.

Tragic and Happy Endings

The first qualification concerns what is meant by
satisfactory. A drama may end tragically as well as
happily. (It also may end tragically in some respects,
happily in others.) Both kinds of ending are satisfactory
to the audience of a drama (everything is settled);
however, a tragic ending is unlikely to mean
satisfactory completion of a commanders mission.

Chapter 5


We are really using the word satisfactory in a scientific

and aesthetic sense. Whether the ending is happy
or tragic, there is stability in the form of contentment
(i.e., no character has anything left to hope for or to
fear). In either case, this comes about partly through
modification of initial hopes and fears during the
course of the drama. At a tragic ending, broadly
speaking, hopes have been destroyed and fears
realized. At a happy ending, hopes have been
realized and fears banished. Clearly, a commander
wants stability at a happy ending, although perhaps
only after hopes have been appropriately modified,
rather than at a tragic one.
For an example of the tragic type of stable ending,
consider the following generalized model of how
countries escalate their differences until they gladly
go to war. Let each implicitly or explicitly threaten to
fight unless its political and diplomatic position is
accepted. Let each then eliminate its inducement
dilemma by rationalizing a preference for war as
compared to the others position. Then, to eliminate
any possibility of the other responding to this by further
escalation, let it take its rationalization so far that war
itself becomes its position, any compromise with the
other being considered worse than war. War has then
become a totally satisfactory resolution; that is, it is
now a shared position from which no group of players
is tempted to defect. Pursue this reasoning further and
we find players escalating the type of war they consider
better than any of the opponents positions until they
are committed to Clausewitzian total war with its
absolutist demand for unconditional surrender.


Confrontation Analysis

The process in which friends and neighbors from

different ethnic groups start to murder each other, as
in Bosnia and Rwanda, may be explained as another
kind of tragic ending. Each group knows, from tales
handed down, that the group its neighbors belong to
has in the past massacred them. Each group hears
officially denied rumors that this group is arming to
give itself the capability of doing it again; therefore,
each group takes the precaution of arming itself.
Finding now that its fears are confirmed, each group
fears that the other now prefers to massacre it, and
hence rationalizes a change in its attitudes so that it
prefers to massacre the other first. The future, Each
tries to massacre the other before it can be massacred
itself, is now necessarily interpreted by any unit within
a group as meaning We succeed in massacring them
first. Unless and until this unit is massacred, this is
the logical way for it to interpret this future. Next, the
rationalizations each group has used to justify its
preference for massacring the other, based on
handed-down stereotypes, means that it now prefers
this to any compromise or settlement. Hence, Each
tries to massacre the other has moved from being
almost the worst outcome for either to becoming a
tragic totally satisfactory resolution.
In general, escalation to a totally tragic ending happens
in the following way. Each party has an inducement
dilemma, which it solves by rationalizing a preference
for the threatened future as compared to the others
position. Each party then has a deterrence dilemma,
which it solves by demonizing the other and thinking
up credible cards that make the threatened future
worse for the other. Each then has an inducement
dilemma, which it solves, and so on. This cycle of

Chapter 5


rationalizations continues until Each doing the worst

it can to the other becomes a shared position elevated
by each above any possible compromise.
To break this cycle, each side needs to start solving
its inducement dilemma not by preferring conflict to
compromise, but by suggesting modifications of its
own or the others position to create a single position
both can share. This is what happens when
negotiations proper start (e.g., during the 3-week
conference in Dayton, OH, that led to the Bosnia
accords). Parties thoughts are then directed toward
agreeing on a single position, although they may not
succeed. Escalation, on the other hand, tends to go
through its successive stages when the parties are
separated, lobbing pronouncements and symbolic
actions at each other from a distance while
addressing either their own constituents or interested
third parties. It is when parties are not formally
negotiating that there is need for escalation to be
controlled and thought given to the construction of a
single positive position. Analyzing how to do this may
be the main contribution of confrontation analysis.

Dealing with the Details

A second qualification to bear in mind in interpreting
totally satisfactory resolution is that the degree of
resolution of a conflict is relative to the model we are
using. Zooming in on a model to see more detail
uncovers potential disagreements that disappear when
we zoom out.
This is literally true of boundary disputes: the more we
zoom in on an agreed boundary, the more potential
disagreements are revealed about where exactly it


Confrontation Analysis

should run. It is also true of agreements on matters

such as cessation of violence, economic arrangements,
prisoner exchanges, and so on.
With this in mind, reconsider figure 2, where characters
cycle between the Buildup and Climax phases until
either they fall into conflict or reach a single, totally
trustworthy position. If the latter, they look at the details
(in the Resolution phase) to see if their agreement is
really stable (i.e., do they mean the same thing and
can they trust each other). If the answer is yes, they
proceed to the Implementation phase.
This is correct in theory; but in applications, is it realistic
to expect a single, totally trustworthy position?
Imperfect agreements reached with much distrust
remaining seem more likely. Our answer is that
imperfect, mistrustful elements of the agreement are
hidden by zooming out. This is realistic. Parties, to
reach agreement, deliberately use ambiguous or overgeneral formulations; therefore, we model them with
card-tables where much detail is covered over by a
few, general cards until all that is left is a single, totally
trustworthy position. That reflects the methods of reallife negotiators.
Nevertheless, you must look into the detail you have
covered up. That is, what parties do in the Resolution
phase. They do it to check that the overall
understanding they have reached is sufficiently sound
and reliable, not to settle all details, which would be a
never-ending task. Card-table models can be used to
see the details of an agreement, as well as its overall
shape. This is accomplished in the following steps:

Chapter 5


Start with a simplified model, representing

parties general positions, to deduce conclusions
about dilemmas and dilemma-elimination.
Subject your conclusions to exploration and
criticism by adding cards and characters to the
model, particularly any that seem likely to
overthrow the conclusions. Assess their effect on
how parties, at the moment of truth, may sum up
issues in the complicated model in terms of
another simple model, perhaps different from the
first. A principle tool for such assessment will be
analysis of subconfrontations between
subcharacters. From your new simple model,
deduce new conclusions about dilemmas and
Criticize your new conclusions again by adding
details to the simple model. Continue until you
have a satisfactory model.
Technically, the technique of card-table modeling can
incorporate any number of cards and characters. The
reason for using a simple model is not practicality; it is
realism. Simple models are realistic because the
characters themselves, at a moment of truth, must
and will use a simple common reference frame to feel
sure they understand each other. To model the frame
in terms of which they reach agreement by, a complex
model would be unrealistic. Complex models, as
complex as possible, nevertheless should be built.
They are built to model not the moment of truth, but
the process by which characters look into the
adequacy of their simple models.


Confrontation Analysis

Example of a Detailed Model (Northern Ireland,

1993)Context Cards
To illustrate, table 7 shows a complex model used to
represent and explore the position supposedly taken
by the British and Irish governments in the Downing
Street Declaration of 1993.
The model not only represents the cards that, under
our interpretation of the two governments position,
should be played by the various parties in
negotiations, it also shows a number of cards that
(under the position) it was assumed would be played
by external characters. These form part of the
assumed environment of the negotiations, and are
placed below the others in a sectioned-off part of the
table headed Context.
In this Context section, the symbol ~ is put against
cards about which either assumption (will be played
or will not be played) could be made. Above the
context line, the same symbol is put against cards
where the Downing Street declaration took no position.
Note that filling in and interpreting a table like this
requires detailed knowledge. This is useful; but it also
underlines the limitations of a complex model. It is
useful in assessing the impact of the details that
underlie general statements and raising questions that
might be important. We can take a comprehensive
look at a mass of detail before summing it up in a
simple model.
It underlines the models limitations by showing why a
complex card-table will be an unrealistic model of a
common reference frame. Characters cannot assume
that such details are common knowledge between

Chapter 5

Table 7. Northern Ireland options in 1993-1994, showing in detail the position

taken by the British and Irish governments in the Downing Street Declaration.


Confrontation Analysis

them, hence they cannot put them into a model of

what they are sure each other knows. They know that
each others knowledge of details varies, with each
appreciating the details of its own situation much better
than the others.

Implementing an Agreementthe Role of

Subcharacters Within a Character
How then are details settled that have been glossed
over by a simple, overall agreement? Each character
consists, in principle, of an organization, and each of
the chiefs that have accepted the agreement must
order it to be implemented within their organization.
This is another reason why the overall agreement
cannot include much detail. High-level officials cannot
decide all the details of a complex operation; they must
devolve them. Hence the orders given to implement
an agreement, and so the agreement itself, must be
simple and general.
Do lower-level decision-makers then simply fill in the
details in accordance with their local knowledge? Not
exactly. What happens is that the resolution of a
conflict generally gives rise to a set of subsidiary
conflicts, as it is implemented through confrontations
between lower-level characters within each
organization. Some of these confrontations are within
organizations, some between them. When we model
the overall agreement, we must realistically try to
assess how such sequences of lower-level
confrontations will work out in practice, just as the
characters will try to assess the same thing at a
moment of truth.

Chapter 5


For example, in 1993 Serbian President Slobodan

Milosevic, a player on the Serbian side who at that
time favored acceptance of the VanceOwen plan,
recommended it to the Bosnian Serbs on the grounds
that it could not actually be implemented. It would leave
Serbian gains more or less untouched. However, this
view was not generally accepted on the Serbian side,
so it did not effectively become the Serbian view. The
incident shows both how characters attempt to
realistically assess what different agreements would
mean in practice and how a characters views arise
out of confrontations between subcharacters, here
Milosevic and the Bosnian Serbs. (Silber and Little,
1996, p. 279).
Often an agreement looks forward to the conflicts that
will arise in its implementation by setting up a
subsidiary organization to help by handing out
rewards and penalties, as was done following the
1995 Dayton accords.
There may be problems when a character in
negotiations has insufficient power to order the
implementation of what it has agreed. In modeling
terms this is not a problem. It is simply a matter of
correctly describing the cards given to characters. If a
character cannot bring about a cease-fire, it should
not have a Cease fire card. Its card should be called
something else, perhaps Recommend cease fire.
Names should indicate characters actual powers. This
again reflects how characters themselves will model
the situation at a moment of truth.


Confrontation Analysis

Unforeseen Contingencies
A third qualification to bear in mind when describing a
resolution as totally stable is that it only reflects
characters projections of a possible future, to which
the actual future is unlikely to conform.
Partly this is because of the previous point. Plans must
be realized through sequences of confrontations
between lower-level subcharacters, and the results of
those confrontations, whether resolutions, conflicts,
or defections, cannot be foreseen in detail. Their total
effect may make implementation of an agreement
better or worse than foreseen.
We do not know the future, so the plans we agree on
are not generally fulfilled, even when we honestly try
to fulfill them and all lower-level conflicts are resolved.
Unforeseen contingencies arise, requiring the
agreement to be interpreted to apply in circumstances
it never envisaged, so that it may need renegotiating
at various levels. This provides yet another reason for
an overall agreement to be simple and general, as it
must provide a framework within which such
adaptations can be made. A clear yet general
agreement allows flexible responses to changing
circumstances while assuring subcharacters that the
agreement remains in force, so that if they implement
it they can continue to expect cooperation from each
other and support from higher authorities. Agreements
that go too far into detail run the risk of being seen too
obviously not to apply when circumstances change.
For example, the history of the United States
Constitution, or other constitutions, shows how a
simple, general agreement is clung to, being

Chapter 5


reinterpreted when necessary, because the framework

it provides for subcharacters activities is too important
and valuable to lose.
We have said before that an agreement that must
survive under circumstances it did not foresee needs
to be entered into in a positive spirit of goodwill, to
reassure each subcharacter that others will strive to
reinterpret and renegotiate it in ways that respect
their interests.

When Conflict Resolution

Breaks Down
The conflict resolution process by nature may always
break down. It can fall into the Conflict phase of figure 2,
the phase where parties prepare to implement their
fallback positions, having failed to bring about sufficient
change to enable them to move from the Climax phase
back to the Buildup phase.
In peace support operations, conflict normally means
resort to or continuation of armed violence. Such
violence may not involve our own force. Others may
do the fighting, such as when UN peacekeepers
withdraw from a deteriorating situation (cf. UNEFs
1967 withdrawal from the Middle East). Armed
violence normally results. In other kinds of
confrontation, such as between different components
or coalition partners on our own side, conflict normally
means a standoff with failure to cooperate, causing a
diminution in operational effectiveness (e.g., failure to
share intelligence or meet minimum interoperability
requirements leading to loss of life through friendly-


Confrontation Analysis

fire incidents, equipment malfunction, or inability to

give timely support).
In disaster relief or humanitarian operations,
breakdown of conflict resolution may mean similar
standoffs, reducing operational effectiveness through
failure to cooperate.

The Details of Implementing a Threatened Future

When parties in the Conflict phase, having failed to
make their threats sufficiently credible or awesome to
deter others, are faced with having to carry them out,
the emotions aroused may be strong enough to
convert the threatened future itself into a formal
resolution of the conflict, as in the case of the tragic
ending discussed above. Such a tragic ending is a
complete, dilemma-less, totally satisfactory resolution;
however, we do not say formally that it goes through
the Resolution phase of figure 2. We say it goes
through the Conflict phase.
We say this to maintain an essential difference
between the Resolution and Conflict phases of figure
2. Characters in the Resolution phase sit down
together to examine the details of their shared
position. In the Conflict phase, each character meets
separately to discuss within itself (i.e., each holding
discussions between its own subcharacters) the
implementation of its fallback position. The characters
at the level of the confrontation as a whole do not
communicate. (Their attempts to communicate would
be defined as attempts to go back to the Climax phase
and restart negotiations.)

Chapter 5


Can the characters in a three-or-more person

confrontation get together and communicate in
This too can be outlawed as a matter of definition.
Characters must have a common view, as part of their
common reference frame, regarding which subgroups
of characters would get together if the threatened
future goes into the Conflict phase. That common view
should determine the definitions of characters and
subcharacters (i.e., we should define communicating
groups of characters as subcharacters of a single
All this is a matter of formal definitions and modeling
procedures, but it reflects important real-world
For example, after the breakdown of the VanceOwen
initiative, the United States and European countries
conferred as to the next step they should take, finally
deciding not to implement their fallback position (lift
and strike); likewise, the Bosnian Serbs conferred with
the Serbian government. The subconfrontation
between them entered its own Conflict phase as
Milosevic angrily cut off military supplies to the Bosnian
Serbs, although afterwards he resumed supplies. This
means that in analyzing the climax of the 1993
confrontation over VanceOwen, we should regard
the Serbs as one character, comprising the Bosnian
Serbs and Serbian government as subcharacters, and
the West as another character comprising the U.S.
and European governments. This reflects the way the
parties perceived which parties they belonged to
(Silber and Little, 1996, p. 287).


Confrontation Analysis

The Conflict Phase Is Game-Theoretic

From this we get a useful theoretical insight. The
decision problem facing characters in the Conflict
phase, after they have absorbed the preference and
belief-changing impact of emotions and
rationalizations generated at the Climax phase, and
provided they do not decide to go back to the Climax
phase, is essentially game-theoretic. All the methods
of present-day game theory (which has, in the main,
turned its back on the tendency to explore and pay
attention to dilemmas) are applicable in principle at
this stage.
There are some qualifications to this. The internal
confrontations between subcharacters are not at all
game-theoretic. Moreover, characters may look
forward to communications in later confrontations with
parties they are now cut off from by the breakdown of
negotiations, and their preferences in their current
situation may be influenced by the positions they
foresee in such later confrontations.
For now they must make a decision in the manner
assumed by game theorists (i.e., without
communication, and hence by choosing the best
course of action within a fixed, given framework of
beliefs and preferences). They must do this whether
or not the threatened future has become a tragic
but nevertheless final resolution caused by emotions
generated in the climax (i.e., whether or not they
still face dilemmas in their relations with other
characters). If they do, the situation requires that
they ignore them and simply pursue the dictates of
instrumental rationality.

Chapter 5


The traditional Clausewitzian military mission is best

understood in this sense. Rapoport (1968, pp. 69-77)
points this out, while nevertheless emphasizing the
limitations of traditional quantitative game theory as a
decision-support tool. Clausewitz assumes that military
specialists are given a mission to carry out against
enemy forces if political negotiations have broken
down and changed national attitudes have brought
about passionate hatred between nations. The
missions objectives are political in that its aim is to
use force to create a situation where resumed political
negotiations will be more favorable to the nation. Its
conduct is non-political. During this military interlude,
negotiations with the enemy concerning the issues
being fought over are out of the question, although
tactical-level negotiations may be carried on over such
issues as cease-fires to remove bodies, local
surrenders, and declarations of open cities.
Meanwhile, relations between the different players on
our side, including the government, various
components of the military, and our allies, are not at
all game-theoretic. Confrontations between their
differing views are supposed to lead to willing
cooperation in meeting the wars objectives.

The Fog of War

Because the Conflict phase is game-theoretic, players
in this phase encounter the fog of war in Clausewitzs
sense. This is because Clausewitzs fog of war has
two causes not present in negotiations when we are
trying to guess at other parties true attitudes and
beliefs to pressure them into accepting our position.


Confrontation Analysis

The first cause of the fog of war, absent during

negotiations, is that physical actions and operations
are being undertaken for their physical effects;
therefore, they are undertaken secretly. If the reason
for undertaking a physical action (e.g., a
bombardment) is to send a message, you will not try
to make it secret. If you do, the message may not be
received. When the aim of the action lies in its
destructive physical effects, keeping it secret has the
advantage of preempting countermeasures. Enemy
secrecy causes fog.
In addition, physical operations tend to meet
unforeseen contingencies, causing them to unfold in
ways we did not expect. This creates more fog as units
ideas of what is happening to each other diverge.
The second main cause of fog in the Conflict phase is
that, because we are not negotiating, we have no
messages to interpret. Someone sending you a
message is trying to make you understand something,
and revealing themselves. There is a problem of what
to believe or not believe. That problem may be called
the fog of confrontation. It is different in kind from the
fog of war, where the problem is one of having no
messages to interpret.

Flunking and Defecting

We have discussed the Resolution phase and the
Conflict phase of figure 2 as if they are followed by
implementation, respectively, of the current common
position and threatened future.

Chapter 5


They may be. But we have said these are not the only
two possibilities. A third is interruption, caused by new
exogenous information overturning the characters
common reference frame (e.g., Pearl Harbor upsetting
the frame of AngloUnited States negotiations). Apart
from this, two other possibilities are the betrayed
resolution and the flunked conflict. The first consists
of one or more characters defecting from the common
position (e.g., the Nazis betrayal of the NaziSoviet
pact). The second consists of one or more defecting
from the threatened future (e.g., the Wests failure to
punish Serbian rejection of the VanceOwen plan).
There are, therefore, five ways a confrontation may
end, usually setting the scene for new ones. Of these,
the decision to flunk is, we have said, game-theoretic
(i.e., instrumentally rational). The preceding Climax
phase may have radically altered our attitude. Also,
we may look forward to future confrontations in which
our credibility may be affected by the actions we take
now. However, given these considerations, we flunk
because we consider flunking to be the best thing for
us in light of our predictions of what others will do.
Consider the classic case of a terrorist who has
threatened to blow up the plane he is on if his demands
are not met. They are not. It is put-up-or-shut-up time.
He now must ask himself, given the emotional
confrontation he has been through, and thinking of
the future he must look forward to if he gives in,
whether he really prefers death. Given his operative
values and beliefs at the time, the decision he makes
will be instrumentally rational.
Is the decision to defect from an agreed position also
game-theoretic? In principle, yes. Having gone through


Confrontation Analysis

the Resolution phase where the parties confirm their

agreement with each other, each must then separately
decide whether to keep its part of the agreement.
Logically, this must be so. There must be some point
at which parties decide whether to keep to their
agreement, otherwise there would never be any
doubt about this, as there clearly is. Yet the gametheoretic nature of this decision is not so clear as in
the parallel case of the Conflict phase. For this there
are several reasons.
In the Resolution phase, parties generally have used
positive emotion and common-interest argumentation
to make themselves prefer to keep to the agreement
rather than pursue any temptation to defect. In this
they may have been successful, particularly if the
discussions have involved subcharacters within each
main character. The result is that when they make
their separate, game-theoretic decisions as to whether
to defect, they often find they now prefer not to.
Secondly, during the Resolution phase parties may
have agreed to an implementation plan instituting
incentives for lower-level decision makers to stick to
the agreement; they will have done so in order to
eliminate temptations to defect. (For example, the
Dayton accords invited into the area the NATO
Implementation Force.) By affecting the incentives
of subcharacters within each main character, this
again makes the main characters prefer, on the
whole, to stick to the agreement.
Thirdly, parties who have reached an agreement often
look forward to a relationship where they will benefit
from cooperating with each other in future

Chapter 5


confrontations; therefore, it is important for them to build

up and maintain positive credibility (i.e., the belief that
they will not defect from agreements). This gives them
a reason not to defect from the present agreement.
To see the importance of the last two factors, compare
two different scenarios in which a British colonel gets
local militia forces to agree to remove its weapons
from a certain area. In Scenario 1 the British
commander will, immediately after the agreement,
withdraw his battalion to be replaced by forces of
another nationality (e.g., Russian), with which British
forces have poor liaison, and which the local militia
think are likely to be more sympathetic to their side.
In Scenario 2, the British battalion stays to oversee
the keeping of the agreement. We may suppose that
under Scenario 1, the British colonel and local militia
commander, knowing what will happen next, go
through a Resolution phase in which they expend
much effort and goodwill convincing each other their
agreement will be kept; however, no matter what they
say, it is more likely to be kept under Scenario 2.

Summary of Chapter 5
When all dilemmas have been eliminated, it is
necessarily the case that all characters agree on a
single position and can be trusted to implement it
(see Appendix). In this sense, a totally satisfactory
resolution has been reached, subject to a number of
provisos. First, satisfactory is meant in a scientific
or aesthetic sense, even though in this sense a tragic
ending with hopes destroyed is as satisfactory as a
happy one in which hopes are fulfilled. In real life, we
generally prefer happy endings. These are brought


Confrontation Analysis

about by managing the resolution process so that

characters attempt to overcome inducement
dilemmas by negotiating a single position, rather than
by escalation. Secondly, resolution is modeldependent. A more detailed model, such as is
examined by the characters in the Resolution phase,
may reveal disagreements that require renegotiation
of what has been agreed. The card-table modeling
technique may be used to examine the details of an
agreement. Thirdly, most details are settled, not in the
agreement itself, but in the course of confrontations
between subcharacters during implementation,
confrontations that may or may not be resolved
satisfactorily. Fourthly, we cannot know the future, so
implementation may turn out unsatisfactorily because
of contingencies not foreseen in the agreement.
Finally, a perfectly satisfactory resolution may be
reached in terms of the common reference frame and
characters communications, with one or more
characters nevertheless intending or later deciding to
defect from it, after having deceived the others into
thinking they were trustworthy.
The same considerations apply, in a different way, if
the confrontation ends not in agreement, but in conflict
(which for defense forces usually means armed
conflict). In the Conflict phase, characters separately
look into details and confer between their
subcharacters as to how to implement their fallback
positions. Detailed consideration may cause them to
change their mind, and either return to the Climax
phase or simply not carry out their threats. When they
do attempt to implement the threatened future, it
usually turns out unexpectedly, leading them into fresh

Chapter 6
The Front-Line Play:
A Dramatization of a
Confrontation Analysis

ost of this chapter consists of a play. It was written

by the author under a contract with the United
Kingdom Defence Evaluation and Research Agency
(DERA). It was performed at a seminar on
confrontation analysis held at DERA (Portsdown
West) on February 5, 1998. Its object was to convey
as vividly as possible how a commander tasked with
a peace mission, and knowing nothing about
confrontation analysis, might use the method to
formulate and begin to implement a strategy for
winning an Operation Other Than War (OOTW). It is
reprinted here with the permission of DERA to give
the reader an example of the formulation and
implementation of a confrontation strategy.

A fictionalized peace mission is used, although with

obvious resemblance to real ones, to focus attention
on the process of analysis and strategy formulation
rather than on the details of a particular case. The
characters are British because the underlying research
was done with British defense forces.



Confrontation Analysis

Would It Be Done This Way?

In the play General Deloitte, commanding a combined
joint task force, analyzes his situation by calling in a
confrontation analyst on his staff. The general knows
nothing about the method; it is something he has just
heard about. He analyses the immediate problem at
his level, then asks the analyst to further analyse the
following items:
Grand strategic problem that gave rise to his
being tasked with this mission
Other problems at his own level, both
simultaneous with the immediate one and to be
expected as the campaign develops (i.e., the
linked sequence of confrontations)
Internal problems of coordination between different
forces in his coalition and different components
Problems at lower levels, consisting of tactical
confrontations that should be resolved in ways
that support and are supported by his
operational strategy.
What can be said about the organizational
arrangements under which the general does this?
Although better than doing no confrontation analysis,
they are imperfect. The general and his chief of staff
should have learned the method in normal training
procedures. They would not then depend on the analyst
for guidance at every step, as they do in the play.
As he does in the play, the analyst would depend on
them for the strategic decisions that would underpin
his more detailed, lower-level models. For these

Chapter 6


models he would need more staff than himself, and

the models would need to be developed through
interactions with responsible commanders at all levels.
In the next chapter we explore the principles of
constructing a confrontation strategy. While reading
the play, remember that here we are showing the
fundamentals of the method by taking the case of a
commander who has encountered it for the first time
and is learning it at the same time he is using it. This
is not meant to be ideal.

The Play
[The scene is a partitioned area of the banqueting hall
of a tourist hotel near Morubwe, the capital of the North
African state of Morya. The hotel has been
requisitioned for HQ UNFORMOR (United Nations
Force in Morya). The particular area we are in is the
office of the commander (UNFORMOR), MajorGeneral Eric Deloitte, CBE. It contains the generals
desk, a large hotel dining table with chairs set round
it, and a flipchart.]
[Commander UNFORMOR is British because Britain
has been designated the framework nation for the UNs
intervention in Morya. Under him is a British brigadier
commanding the Anglo-Egyptian land forces, a U.S.
Air Force brigadier-general commanding the air forces,
and a colonel of the Arme Francaise commanding
the aviation force, a battalion of helicopters available
to support the land forces.]
[General Deloitte, a determined-looking, middle-aged
man in combat dress, is standing looking discontently
out of the window. Seated at the table is his chief of


Confrontation Analysis

staff, Brigadier Ray Jones. The chairs are drawn back

from the table and papers are scattered on it. Evidently
a meeting has just finished. On the flipchart are
scrawled the names of some of the main protagonists
in the Moryan drama]:
border, supplies ISRA)
FRANCE (tends to support Moryan govt)
EGYPT (strongly against ISRA)
ARAB COUNTRIES (some support for ISRA)
[Brigadier Jones is busy at the table sorting through
papers and making notes.]
DELOITTE: Is everything ready for the press
JONES [still writing notes]: Yes sir. At the airport.
Fourteen hundred. We should leave 1330.
DELOITTE [irritably]: Whats the right way to handle
this? Its no good, you know, Ray. Were not there.
JONES [raising his eyebrows and turning to look at
the Generals back.]: Really, sir? This mornings
meeting was the most thorough review of the situation
weve had. Im just writing up the notes for you. Each
component reviewed the situation from their viewpoint.
Intelligence gave a good overall assessment. We know
with fair accuracy whats going on, how the relief
convoys are getting through, etc. As to the press,

Chapter 6


Ive seen you handle them. Im sure you can refrain

from giving away anything thatll further inflame the
DELOITTE [disgustedly]: That, precisely, is not the
point. This press conference should be my first move
in a planned campaign to reach my objectives. It
shouldnt be a matter of giving nothing away, of giving
no hostages to fortune. But Ive nothing resembling a
plan. [Turning to face Jones] Thats the trouble, Ray.
Soldiers know how to fight battles. I can plan for a
battle. But this, this is not anything resembling a battle.
The objective here is to get your way without the use
of force. We are not trained to do that. Thats a
politicians job.
JONES [wisely]: Peacekeeping isnt a soldiers job,
but only a soldier can do it.
DELOITTE: Who said that? Never mind, its quite
right [Goes back to gazing morosely out of the
window. Then starts again, waving his hand at the
flipchart.] You know, Ray, Ive a hunch the Moryan
government, not ISRA, will be the one that refuses to
cooperate in peace talks. Apparently we still have not
got the President to agree to a time and place for an
emergency meeting with ISRA. Thats why they called
me out of the meeting half an hour ago. I told them to
keep trying.
JONES [surprised]: The government make difficulties?
Why should they? Our intelligence is that theyre under
the most pressure to end the fighting. They have most
to lose from it continuing.
DELOITTE: That was the position. My hunch is that
things have changed. I suspect they think all this world


Confrontation Analysis

publicity against ISRA has tipped the balance in their

favor. They think itll force us to intervene on their side.
So, the worse things get for them the better. That, I
suspect, is their thinking.
[Jones shakes his head disgustedly.]
DELOITTE: Meanwhile, ISRA will continue their
campaign, regardless what the world thinks. And Im
supposed to do something about it. Without using
force. But what? [Musingly] You know, Im trying to
remember something I heard recently. About a new
thing called confrontation analysis. Thats what I need.
After all, thats what PJHQ has thrown me into. A
confrontation. Thats the term for it.
JONES [hesitantly]: Sir
JONES: Weve someone on the HQ staff whos trained
in confrontation analysis. He is a civilian analyst in
OA called Bright, Mark Bright. DERA sent him along.
We havent used him in that capacity. OA have had
their hands full programming databases for us. But I
could get him in to tell you about it if you like.
[The Commander looks at Jones thoughtfully, as if
his mind were on other things. Then he shakes his
head as if to reproach himself for indecision, and
glances at his watch.]
DELOITTE: Send him in.
JONES: Yes sir.
[Jones gets smartly to his feet and leaves the room.
The general returns to his morose vigil at the window.

Chapter 6


Almost at once Brigadier Jones returns leading Mark

Bright, an intense and slightly supercilious young man
in civvies. The general turns round. Bright, without
actually saluting, stiffens and inclines his head as if to
indicate military readiness.]
JONES: This is Dr. Bright, sir.
DELOITTE [regarding him curiously and somewhat
humorously]: Sit down, doctor. [Bright does so.] Tell
me about this new discipline DERA has trained you in.
BRIGHT [promptly]: Yes sir. Confrontation analysis
is an OA tool. You analyze situations where each party
takes a position, meaning a suggested solution, what
it suggests everyone should do, itself included.
Generally, parties positions dont coincide. Theyre
proposing different solutions. So theres a conflict of
wills. As well as their positive positions, each party in
a confrontation has a fallback position, what it says
itll do if its position is not accepted.
DELOITTE: That defines a confrontation?
BRIGHT: Yes sir. In a way confrontations correspond
to battles. To win a war, you win a linked sequence of
battles. To win a peace support operation, you could
say, you fight and win a linked sequence of
DELOITTE [in a lively tone]: Well, suppose Im in a
confrontation. How do I analyze it?
BRIGHT: Well, once youve decided on everyones
positions, you look for the so-called dilemmas facing
each party. Essentially, dilemmas are credibility
problems. For example, I have a threat dilemma if you
think I would not carry out my threat. In other words, if


Confrontation Analysis

you think Im bluffing. I have a trust dilemma if I couldnt

trust you to carry out my position, even if you agreed
to it. There are four other dilemmas.
DELOITTE: They sound realistic enough.
BRIGHT: Theyre extremely realistic, in my
experience, sir.
DELOITTE: Can you always find these dilemmas? In
any confrontation you look at?
BRIGHT: Theyre guaranteed to exist, sir, unless and
until all parties have agreed on a single solution they
can trust each other to carry out. In other words, if
theres no dilemma, theres no problem. If theres a
problem, the parties are necessarily facing dilemmas.
So by eliminating all dilemmas, you solve the problem.
DELOITTE [bridling]: You do, do you? On whose
terms? Whose solution do you end up supporting?
BRIGHT: We have to make sure its our solution that
wins, dont we, sir?
DELOITTE [loudly]: And how the hell do we do that?
[He leans forward with his hands wide apart on the
table looking down at Bright, who sits across from him.]
BRIGHT [quietly, looking down at the table]: Its a
matter of making a plan and following it. Winning a
confrontation is like winning a battle, only different.
The details are different, but the principles the same.
You exploit the other sides weaknesses and our own
strengths. First off, you analyze the dilemmas
everyone faces. Dilemmas are points where we can
exert pressure on them, or, if were not careful, them
on us. Next, you make a plan to eliminate dilemmas

Chapter 6


in a sequence of operations that will bring the others

into full compliance with your objectives. [He looks up
to find the general staring fixedly at him. He pauses
uncertainly, then continues more quietly than before.]
Its a logical process, sir. Supposedly, its as logical,
or more so, than winning a battle. Of course, youve
still got the fog of war to contend with. I mean, you
have to make assumptions because you cant be sure
of the facts, and your analysis is dependent on the
facts you put in. But youve got to act. So you make
the best assumptions you can, whilst being ready to
change your plan, without vacillation or confusion, if
your assumptions prove to be wrong [His voice
falters as the general continues to glower at him.]
DELOITTE: Youre quoting Clausewitz.
DELOITTE: Thats straight Clausewitz.
BRIGHT: Yes sir.
DELOITTE: What does winning mean when youre in
a confrontation?
BRIGHT [without hesitation]: Bringing others into full,
willing compliance with your objectives.
DELOITTE [turns his back and walks to look out of
the window. After a pause]: Okay, suppose I want to
analyze the confrontation Im in now. How do I do it?
What do I need? [Looking at his watch.] How much
can I do in one hour?
[Bright looks appealingly at Jones, who returns his
look impassively. The general continues to gaze out
of the window, his back to them.]


Confrontation Analysis

BRIGHT: Er, well, sir, I could take you through a broad,

high-level analysis. Id sort of ask you a series of
questions and show what follows from the answers.
Youd be responsible for all assumptions, and
obviously decisions taken on the basis of the analysis
would be yours. Id be responsible for the process of
analysis, showing you how to do itstep by step
JONES [imperturbably]: Equipment needed?
BRIGHT: Just a flipchart. In an hour, we mightmap
out a broad policy, though probably itd raise a lot of
questions for intelligence to answer. Wed have to do
more detailed analysis and checking later on. Then
we could make detailed plans to be implemented at
various levels and by various components. [Looking
anxiously at Jones while still talking to the general,
whose back is still turned to them] Id have to start,
sir, by asking what decisions have to be made or
actions taken in an hours time.
[A long pause.]
DELOITTE [turning round briskly and coming to sit at
the table]: Of course. [Waving his hand at the flipchart]:
Theres your equipment. Ray, perhaps youd better
sit in on this. I may need your views on some of
theerassumptions. [To Bright]: Lets go.
BRIGHT [standing up awkwardly and going to the
flipchart]: Yes sir. Er
DELOITTE [sitting easily, his hands wide apart on the
table]: To answer your first question. I have to give a
press conference at 1400 hours. I could just regard it
as a nuisance, and give away as little as possible.
That would be our usual style, I suppose. [He looks

Chapter 6


mischievously at Brigadier Jones, who does not

respond.] Instead, I want to handle it as the first step
in a PLANto achieve my OBJECTIVES. So, at 1400
Ill be facing reporters and cameramen from all the
worlds media. These are the people whove whipped
up public hysteria in Britain, America, and Europe over
the Friday massacre attributed to ISRA. Dead babies
on television. All that kind of thing. Horrified the whole
world. The Cabinet met to discuss it yesterday. The
Defence Secretary contacted the CJO, who called me
in. I got back from PJHQ early this morning.
BRIGHT: And you came back with a new mission, sir?
DELOITTE [sardonically]: My mission, if you can call
it that, is to get the politicians out of trouble. Somehow
or other, I have to stop ISRA massacring people, or at
least, stop it getting into the news. Once again, weve
been landed with what ought to be the politicians job.
But CJO overrode my objections. Said weve got to
do something.
BRIGHT [hesitantly]: So, sir, your objective is? How
would you state your strategic objective?
DELOITTE [after pausing to think]: To get a ceasefire between ISRA and the government, with an
agreement between them to start discussing a
settlement. Thats my objective. To be achieved
without using force.
BRIGHT: If possible
[The general does not answer.]
BRIGHT: I mean, its possible the only way to get
your way without using force is to be ready to use
force, if necessary. Wouldnt you agree?


Confrontation Analysis

[The general gives a short, bitter laugh. Brigadier Jones

smiles wanly.]
DELOITTE: Youre right, in general. But youre
straying into a conceptual minefield. Doctrine, so far
as weve got one, distinguishes peacekeeping, where
you dont use or threaten force, from peace
enforcement, where you do. Were supposed to be
here in a peacekeeping role. On the other hand, a
peacekeeping situation is one where the parties are
prepared to be peaceful. In Morya, theyre not. So
were in a peace enforcement situation with a
peacekeeping mandate. What do we do about that?
BRIGHT [rising to the challenge, picking up a pen and
folding up the flipchart to disclose a blank sheet of
paper]: I cant, of course, tell you what to do, sir. [Jones
raises his eyebrows at this nave remark.] All I can do
is help you analyze the situation to see whats possible,
that is, how far its possible to get peace without
threatening force. And if so, how. Now

First Narrated Interlude

[As he raises his pen, the narrator steps forward and
lifts his hand. The three characters freeze in position.]
NARRATOR: General Deloitte, assisted on request
by Brigadier Jones, now answers Mark Brights
questions as to the positions currently being taken by
the parties. This enables Mark to draw up this cardtable. [He folds up the blank sheet on the flipchart to
disclose a sheet, shown here as table 8.]
NARRATOR: As Mark explains to the general, he
has modeled the confrontation by giving each
protagonist certain cards to play or not play. The

Chapter 6


various parties positions are then shown by listing, in

separate columns, the cards each party proposes
should be played.
So the UNFORMOR position, column U, is that both
sides should cease fire and join peace talks.
UNFORMOR then will not call air strikes against ISRA
(although this is what the Moryan government wants)
nor recommend UN withdrawal (which is what ISRA
wants). Note the name given to the card, Recommend
withdrawal. The general cant actually decide on a

Table 8. Moryan government no longer wants a cease-fire.


Confrontation Analysis

withdrawal, he can only recommend it to his superiors.

Hence the name.
ISRAs position, column I, is that the UN should
withdraw, and therefore that the general should
recommend this to his superiors. He shouldnt call
air strikes against ISRA positions. ISRA and the
Moryan government should be left to fight it out. ISRA
expects to win the resulting civil war and install an
Islamic regime.
The Moryan governments position, column M,
apparently has just changed. It seems unwilling to
attend the meeting the general has called, which
indicates that its now taking the position that there
should be no cease-fire or peace talks. Instead,
UNFORMOR should intervene on its side by calling
air strikes against ISRA. Until now, while continually
asking for air strikes, the government hasnt objected
to the idea of a cease-fire followed by talks.
These are the parties positive positions. But a
confrontation consists not just of positive positions. It
includes parties fallback positions as well, the
unilateral actions they say theyll take if others dont
convincingly agree to their positions. If all parties carry
out their fallback positions, we get something called
the threatened future. This is shown in column t. It so
happens that in this case its the same as the default
future (column d), the future to be expected if everyone
continues their present policies without change. This
is a future under which ISRA and the Moryan
government will continue to fight while UNFORMOR
limits itself to humanitarian aid. So UNFORMOR does
not call air strikes nor recommend withdrawal, while

Chapter 6


ISRA and the Moryan government neither cease fire

nor join talks.
This is the future that has led, through one particular
incident, to the present crisis. The general situation is
that government forces are maintaining control of the
center of Morubwe and other big towns, while ISRA
control most of the country. ISRA are being supplied
militarily by the neighboring government of Pelugya,
which has a radical, fundamentalist Islamic
government, while bombarding government-controlled
areas, causing deaths which tend to get reported by
the Western media. The particular incident which has
now caused an international outcry and led to the
general being told to do something is the bombardment
of a crowded market, which caused sixty deaths.
And the background to all this?
Three years ago ISRA supposedly won an election. But
the secular government of Morya, headed by President
Saldin, hoping for Western support, annulled the
election results. ISRA thereupon began a bloody revolt
and Saldin asked for Western help. The UN approved
the sending of UNFORMOR with a mandate not to take
sides, but to protect humanitarian relief missions while
encouraging the parties to agree on fresh, UNsupervised elections. ISRA have rejected this plan,
arguing that they have already won elections. President
Saldin has gone along with the plan until now in order
keep Western support, but he really wants the West to
intervene militarily and help him defeat ISRA.
Meanwhile, the UN has banned military flights and
imposed sanctions against Pelugya for supplying ISRA,
while UNFORMOR has brought in aid to the populations
of the cities and sent convoys to aid rural areas, entering


Confrontation Analysis

into negotiations with ISRA to allow them passage. Thus

it has built up cooperative relations with both sides in
pursuit of its humanitarian objectives.
[The narrator waves his hand, causing the characters
to come back to life. He walks off the stage and leaves
them to it. The general stands up and leans forward
across the table, gazing intently at the flipchart.]

Resumption of the Play

BRIGHT: So it seems this [indicating the flipchart] is
the moment of truth in the drama between you, ISRA,
and the Moryan government. Obviously, there are
other dramas. Theres the higher-level, Grand
Strategic drama between governmentsthe U.S.,
Britain, France, the Moryan government, Pelugya,
various Arab governments. Thats the drama that led
to the formation of UNFORMOR, that is, its the drama
that defines your superiors intent at two levels above.
DELOITTE [sardonically]: If there is such a thing.
BRIGHT: I mean, sir, that the requirement for a
commander to understand his superiors intent, two
levels up, is not very clear in this kind of UN peace
operation. In this kind of case superiors intent is the
resultant of political interactions between governments.
So understanding that grand strategic drama is
essential. So is understanding the lower-level, tactical
dramas that are also going on between the forces you
command and other forces, as well as NGOs, aid
organizations, communities, and so on. You have to
direct those lower-level dramas to implement a whole,
cohesive plan for achieving your objectives.

Chapter 6


JONES [glancing at his watch]: We dont have time for

all this, Dr. Bright. The general meets the press at 1400.
BRIGHT: I understand that, sir. Analyzing those other
dramas will have to be done later, in follow-up work.
[ Gesturing at the flipchart ] This is the one well
concentrate on now, but we have to remember those
others in the background because well be making
assumptions about them.
DELOITTE [as if talking to himself]: How do I get them
to agree on column U?
BRIGHT: Ah. Yes, sir. First we have to analyze the
dilemmas everyone faces.
DELOITTE [with a quick gesture]: Show me those.
BRIGHT: Ill have to ask you some questions about
players preferences. FirstIsnt it true that both ISRA
and the Moryan government prefer this, the threatened
future (the same as the present, default future) to our
position, column U? Isnt that correct? [As Bright refers
to each future, he points to the corresponding column
on the flipchart.]
DELOITTE: HmmI dont know about that. ISRA
wants us outSaldin wants us to intervene against
BRIGHT: Right. The present future is not the position
of either of them. But in the case of the Moryan
government, surely it hopes, by milking the present
situation, to get the UN to tell us to intervene against
ISRA. Notice, here were talking about its assumptions
concerning the grand strategic drama.


Confrontation Analysis

DELOITTE: In that sense I believe youre right.

Correct, Ray?
[Brigadier Jones nods.]
BRIGHT: So, taking into account those assumptions
about the grand strategic drama, the government does
prefer to continue with the current future, the same as
the threatened future, rather than move with ISRA to
our position. [ They both nod. ] Okay. Now,
ISRASurely ISRA believe the current conflict will end
in their victory.
JONES: Thats why they want us out, Dr. Bright.
BRIGHT: Yes, in other words theyd most of all prefer
this, their position [indicating column I]. But between
continuation of the present conflict with us here but
remaining neutral, and a cease-fire with peace talks
(which is our position), theyd prefer the present conflict.
DELOITTE: Correct.
BRIGHT [triumphantly]: Then we have it. We face a
deterrence dilemma. Fatal. We must solve it, or our
position is untenable. Unrealistic.
DELOITTE [sardonically]: Presumably youre trying
to say that because they prefer being where they are,
theyre under no pressure to accept our position.
BRIGHT: Exactly. Or rather, they prefer what were
threatening them with, which in this case happens to
be the same as the present, default future, to our
position. So, as you say, were placing no pressure
on them to accept our position. U is, at present, simply
out of the question. Thats our deterrence dilemma:
weve got no deterrence.

Chapter 6


DELOITTE: I take your point. I congratulate you, Dr.

Bright, in having gone to the heart of my dilemma. [He
walks up and down, hands behind his back, frowning.]
The position seems to be rather hopeless.
[Jones looks at the general with concern.]
BRIGHT: Erexcuse me. [Jones looks at him sharply.]
May I go through the ways of solving a deterrence
dilemma? [ The general stops walking and looks
inquiringly at him.] One: you can, of course, change
your position. You must do that if all else fails. Two:
you can persuade one of the other parties that the
threatened future, or anything they might do in reaction
to it, actually makes them worse off than our position.
DELOITTE: But weve just said they prefer it.
BRIGHT: But why do they? Take the Moryan
government. They only prefer the threatened future
because theyre hoping for certain grand strategic
reactions, press hysteria, leading to a demand for
intervention against ISRA. Cant we persuade them
that will never happen?
JONES: We cant get into playing politics.
BRIGHT [to the general]: Sir, could we brainstorm
this a bit? Without analyzing the grand strategic drama,
which we havent time for, perhaps theres something
you could do, make recommendations to your political
masters, to get a grand strategic message sent to
President Saldin to discourage him from expecting
intervention. Or is there some way you could use, or
threaten to use, the press conference?
DELOITTE [staring at him]: Of course there bloody
is! Ray, remember what Intelligence was telling us.


Confrontation Analysis

Theres evidence the Moryan government may have

bombed their own people, that this whole thing was
set up by them. That fits in with the way they seem to
have suddenly switched their line against peace talks.
JONES: But sir, it was decided not to use
DELOITTE [belligerently]: Decided? So what? [To
Bright] Lets think about this. If at the press conference
I were to present evidence that the government itself
is to blame for the bombing[Thinking again] But no.
Press reaction might be unfortunate.
BRIGHT: Sir, like most threats, that might be more
effective as a threat than it would if carried out. If I
may, let me give you a card

Second Narrated Interlude

[The narrator steps forward and lifts his hand, freezing
the actors in position. He raises the sheet on the
flipchart, revealing table 9.]
NARRATOR [pointing to rows and columns in table 9
to illustrate what he says]: Mark now adds the row,
Blame the government, to represent a card held by
UNFORMOR. The general then decides to declare
his intention to play this card if theres no movement
in others positions, specifically, if the Moryan president
refuses to talk. In other words, he makes it part of a
new fallback position for UNFORMOR, part of a new
threatened future, shown in column t. Brigadier Jones,
incidentally, points out that ISRA have in fact blamed
the Moryan government for the bombing, which
indicates that for UNFORMOR to blame them will be
part of ISRAs position, too (column I).

Chapter 6


[Turning to the audience]: Now, whats going on here,

looked at from a technical viewpoint? The general has
reacted to his deterrence dilemma, pointed out to him
by Mark, by reframing the situation, adding a card to
his hand, and inserting that card into the threatened
future. This eliminates the dilemma by making the
threatened future worse for the Moryan government
than his (the generals) position. Notice the emotion
with which he carries out this reframing. He becomes
angry at the Moryan government. But he doesnt get
angry on perceiving his dilemma (then he was merely

Table 9. The general has thought up a

new card, Blame the government.


Confrontation Analysis

dejected, walking up and down, fearing he might have

to abandon his position). Nor did he get angry, much
earlier, at the Moryan reluctance to talk. He got angry
when starting to think of a way to make the threatened
future worse for them. Anger is the emotion that
accompanies thinking up an adequate threat.
Having thought up the card, the generals anger
disappears while he considers how to use it. Does he
face any dilemmas in making it credible that he will
use this card? Watch
[He waves his hand again, bringing the actors back to

Second Resumption
BRIGHT [ pointing to the new card]: Blame the
government. If you make that part of the threatened
DELOITTE [ thoughtfully ]: then from President
Saldins point of view, that threatened future would be
pretty bad. With us blaming them for the bombing, how
could they expect the West to intervene on their side?
BRIGHT: Exactly. Adding this card makes the
threatened future worse for them than your position,
and so it eliminates your dilemma. Instead, it gives
the Moryan government a dilemma, an inducement
dilemma, consisting in the fact that they now prefer
our position to the threatened future.
DELOITTE: You call that an inducement dilemma?
BRIGHT: Yes, sir. It puts pressure on them to accept
our position. Its good for us, bad for them. We, on the

Chapter 6


other hand, dont prefer their position to the threatened

future. Weve no inducement dilemmaLets see,
now, what about a threat dilemma
DELOITTE: Hopefully I will not have to actually play
the new cardThe threatened future is one that
neednt actually be implemented
BRIGHT: Correct, sir! Its a threat that lies in the
background, hopefully making them change, because
if theres no change, it represents whats expected to
happenBut, if I may, can we look at other dilemmas?
Is the new card, Blame the government, one youd
want to play, sir? For its own sake, I mean, apart from
putting pressure on the president?
DELOITTE: Er no. Id obviously prefer not to take
sides, upset world opinion, act against the very side
thats favored by our own and allied
governmentsDefinitely not preferred.
BRIGHT: Then, sir, we do face a threat dilemma,
consisting in the need to make credible a threat wed
rather not carry out. How to overcome it? Well,
assuming we dont want to change position, we can
use emotion. Getting angry could make them think
well carry out the threat, even while preferring not
toBetter, can we rationalize a change of preferences,
that is, give reasons why we would actually prefer to
or be forced to carry out the threat? Can we think of
any such reasons?
JONES: May I make a suggestion, sir?


Confrontation Analysis

JONES [delicately]: You might indicate that it would

be your duty, if they refuse to talk, to reveal the
evidence that puts blame on the government. There
is, perhaps, a perception on their part of the British
BRIGHT: Right! I see what you mean! A perception
that the British can be unreasonably obstinate in
pursuit of inexplicable principles! [Jones nods.] That
way, sir, you could simultaneously admit all the
reasons why youd regret revealing this evidence to
the press, while making it clear that youd feel forced
to do so. On principle
DELOITTE [decisively]: Ray, can we get through to
Saldin, on the phone?
JONES: Yes sir. The communications room
DELOITTE: Im going to do this right now, before the
press conference. I can see just how to put this to him.
[He leaves the room. Bright becomes plunged in
thought in front of the flipchart. Jones stands up, walks
in front of the table, folds his arms, and gazes
thoughtfully at Brights back.]
JONES: Dr. Bright, may I ask you something?
BRIGHT [turning round]: Of course.
JONES: Is there anethical aspect to use of these
procedures? It seems somewhatMachiavellian. Not
a straightforward way to deal with people.
BRIGHT: I see. Youre right, in a sense, when were
using the procedures in this way, for decision support
of one party. Then its a matter of helping that party to

Chapter 6


get its own way. We do it by helping it to behave

naturally, that is, transmit emotions and arguments
appropriate to its situation. Thats all right, as far as it
goes. But it also means making its behavior
consciously appropriate and functional, while leaving
all the other parties to stumble around and mess up
their chances. So the situation is asymmetrical. Youre
showing one side how to win.
JONES [nodding thoughtfully]: Which is what we want,
of course
BRIGHT: I assume so. Id like to assume an Allied
peace-keeping mission is generally in the right, and
ought to get its way. In another kind of situation, we
can use the same kind of procedures for mediation
support instead. That means working simultaneously
with all the parties, getting them to solve dilemmas in
such a way as to work out a cooperative solution
[He tails off as the general re-enters the room. Jones
smartens up and turns his attention to his commander.]
DELOITTE: Right. New situation. Assume the
presidents changed his position.

Third Interlude
[The narrator steps forward, raises his hand to freeze
them in position, and lifts another sheet on the flipchart
to reveal table 10.]
NARRATOR: General Deloittes blunt speaking to
President Saldin has shifted the Moryan government
into acceptance of UNFORMORs position. The
general now wants to press home this victory. So Mark


Confrontation Analysis

Table 10. After the Moryan president has

accepted the UNFORMOR position.

sets up a card-table in which UNFORMOR and the

government occupy the same position.
He then asks, what is now UNFORMORs fallback? Not
Blame the government. Thats not necessary any more,
now the government has accepted UNFORMORs
position. Does then the fallback consist of doing nothing?
If so, the general will face another deterrence dilemma,
since ISRA (the only player now openly rejecting his

Chapter 6


position) will prefer such a threatened future to the

generals position.
They reluctantly decide that threatening to call air
strikes is all they have to bring pressure on ISRA. This
is an action the general would be most unwilling to
take. It would jeopardize the relief convoys that depend
on ISRAs cooperation to get through. Also, it might
cause ISRA to retaliate against UN personnelbut
as soon as this possibility is suggested, Mark points
out that it is not in the model. It has to be modeled as
an important ISRA card. He puts it into the card-table
and makes it part of the threatened future, ISRA having
more than once indicated that this is what it would do
if UNFORMOR were to order air strikes.
The discussion now continues
[The narrator waves his hand, bringing the characters
back to life, and leaves the stage.]

Third Resumption
BRIGHT: Okay, so ISRA will threaten us with
retaliation. Fair enough. But is their threat credible?
Havent they got a threat dilemma? Does retaliating
against UN personnel make the threatened future
better for ISRA, in and for itself, disregarding any
pressure it puts on us?
DELOITTE [dubiously]: Hard to say, in and for itself.
That is, forgetting about possible international
BRIGHT: Just a moment, sir, we shouldnt forget about
them. Theyll be part of the grand strategic game,
which, remember, we have not analyzed, but have to


Confrontation Analysis

consider when analyzing this one. International

repercussionsLets seeyou mean that if they
retaliate against UN personnel, its likely the UN and
the powers behind it would demand intervention
against ISRA?
DELOITTE: Precisely. Dont you agree, Ray?
JONES: I do, sir.
BRIGHT: Okay, then if we take grand strategic
repercussions into account, ISRA should have a threat
dilemma in threatening retaliation. If they realize it. I
mean, if they realize the likelihood of repercussions,
they should prefer not to retaliate against UN
personnelIf they dont realize it, perhaps we should
point it out to them! Which suggests something
elseWe might at the same time get rid of our own
threat dilemma, I mean the fact that theyll think us
reluctant to carry out air strikes.
DELOITTE: There youve lost me, Im afraid. Go over
that slowly.
BRIGHT: Yes. I mean this, sir. Weve established, I
think, that heavy, sustained air strikes against ISRA
positions would solve our deterrence dilemma. That
is, rather than suffer that, theyd prefer to accept our
position, which is a cease-fire and peace talks.
JONES: Although weve said their acceptance
mightnt be very trustworthy
BRIGHT: Right, agreed. Im not forgetting that. That
is to say that accepting our position might give ISRA
(and the government, too, come to that) a cooperation
dilemma, in that theyd be accepting a position it pays
them not to honestly implement. Likewise itd give us

Chapter 6


a trust dilemma, wed have got them to agree to

something we cant trust them to do. But those are
dilemmas to be tackled once we get overt acceptance
of our position. Getting overt acceptance is the first
step. Thats what were discussing now.
DELOITTE: One step at a time.
BRIGHT: Exactly, sir. Now, to get overt acceptance
we must make it the best alternative for them that they
can see. That means, first, making the alternative, the
threatened future, worse for them, that is, solving our
deterrence dilemma. Air strikes do that. Correct?
DELOITTE: I believe so.
BRIGHT: Second, it means making air strikes credible.
That means getting rid of our threat dilemma. Making
them believe we prefer to or will be forced to carry out
air strikes, if they dont accept our position.
DELOITTE [concentrating on this]: Even though I
dont, in fact, prefer to carry them out, because, as
weve said, itll mean an end to relief convoys plus a
likelihood of retaliation.
BRIGHT: Correct, sir, but we have to get rid of that
impression in the minds of ISRA. How do we do that?
Well, we could emphasize to them the certainty of
international repercussions if they retaliate. If they
become convinced of that, their retaliation may
become unlikely, eliminating that particular reason for
our dislike of air strikes.
DELOITTE: Okay. But how do we eliminate our other
reason for disliking air strikes? The fact that it will
disrupt relief convoys?


Confrontation Analysis

BRIGHT: Er, maybe Im confused here. I had in mind,

again, emphasizing international repercussions as the
reason why
DELOITTE [exuberantly, completing his sentence for
him]: as the reason why Ill be forced to call air
strikes if they refuse a cease-fire, regardless of whether
I like air strikes or not! Jolly good! This works, Ray. Id
been planning, in the usual kind of way, to dampen
down the tendency toward escalation. Now Im seeing
that tendency as my strength. I mean to use it like a
following wind. Unless we get a cease-fire and talks
pronto, then the pro-intervention movement in the West
will force my hand.
JONES [ dubiously ]: There certainly is such a
movement, sir, pressing for intervention against ISRA.
DELOITTE [walking up and down with hands behind
his back]: Indeed there is. What this has made me
see [waving his hand at the flipchart] is that to avoid
intervention, I shouldnt seem to oppose it. Those
strong pro-intervention forces are what give my threat,
the threat of air strikes, credibility. And without such
credibility, ISRA will not agree to a cease-fire, things
will continue to get worse, until eventually were forced
to intervene on a larger scale and with less chance
of success.
BRIGHT [hurriedly]: Of course, what you say at the
press conference and through other channels may still
fail to make air strikes credible, sir. ISRA may still think
they can call our bluff, given our past history of threats
not followed up. In which case, credibility may require
starting to carry out air strikes, in other words, shifting
the default future to one that contains air strikes.

Chapter 6


DELOITTE [after walking up and down silently for a

while]: Right. So I may have to start air strikes, while
making it clear that if they agree to a cease-fire and
talks, air strikes will cease. We should, therefore, start
withdrawing our people; itll take a while anyhow to
locate targets properly. [He stops, looks at his watch
and addresses Mark.] Dr. Bright, this has been useful.
Im off to the press conference now with a clear overall
plan to pursue. I want you to continue this analysis.
Brigadier Jones will make sure you get the help you
need, including access to informationthe, er,
assumptions you need to put into your model. What
will you look at next?
BRIGHT: Id like to explicitly model both the grand
strategic as well as the military strategic games, sir. I
think we need to be clear as to how they work, how
they impact on your own, operational level. [The
general nods.] Then Ill need to model some of the
confrontations going on at tactical level. We should
plan to send a single, unified message at all levels.
That way we can bring it about that ISRA and the
Moryan government are being pushed in the same
direction by pressures from their own grass-roots as
we are pushing them at the higher level. Thatll be
important for making air strikes credible. Itll be still
more important when it comes to making sure they
keep to our position, that they actually implement it,
as distinct from merely agreeing to it to keep us off
their backs.
DELOITTE [who has been listening intently]: Right.
Carry on. [Turning to Jones] Ray
JONES [opening the door for him]: Yes sir. Your car
must be waiting. Ill see Dr. Bright gets what he needs.


Confrontation Analysis

[They leave together. Mark Bright collapses on a chair

and grins at the audience.]
BRIGHT: Whew! Well, that seemed to go all right.
What did you think?

Summary of Chapter 6
The chapter dramatizes the situation of a commander
who, knowing nothing of confrontation analysis,
formulates and begins to implement a confrontation
strategy with the help of an analyst who happens to
be on his staff. This is not how it would ideally be
done, but serves to illustrate the concept of a
confrontation strategy before a discussion of its
principles in the next chapter.
General Deloitte has been tasked to get two warring
parties to cease fire and start peace talks. He is not to
use force. He has to address a press conference in
an hour, and is anxious to formulate a plan. The analyst
called in to help him draws up a card-table on a
flipchart, points out that he has a deterrence dilemma
and suggests directions in which he might look to try
to eliminate it. This prompts the general to think of a
way forward by adding a new card to his threat against
one of the parties. After discussion of how to remove
his resultant inducement and threat dilemmas, he
implements this plan through a phone call, which
brings the party concerned into apparent compliance
with his mission objectives (i.e., they now say they
accept his position). Of course, this party may have a
cooperation dilemma, so that the general may still have
a trust dilemma; this, however, can be dealt with later.

Chapter 6


A deterrence dilemma remains in relation to the other

party, which, they decide, can only be eliminated by
threatening to use force. A threat of retaliation against
UN personnel can, they decide, itself be deterred. The
generals own threat and inducement dilemmas can
be removed in the eyes of the other party. The general
now has a clear plan upon which to proceed in dealing
with the press conference. He leaves after asking the
analyst to build models of the grand strategic, military
strategic, and tactical levels.

Chapter 7
Formulating A
Confrontation Strategy

he last chapter showed a confrontation strategy

in use, although not as it should be used, not
implemented in an organized way by personnel trained
in how to do it. We saw it adopted in an ad hoc manner
by a commander relying on a single expert who
happened to be on his staff.
Nevertheless, we now have an example to fix on.
Next is a discussion of the principles that should
underlie the formulation and implementation of a
confrontation strategy.

What Is a Confrontation
We have said that Operations Other Than War
(OOTW), and peace operations in particular, can be
viewed as sequences of linked confrontations;
however, the confrontations involved are not all on
one level.
To see what a confrontation strategy is we need to
see how it is implemented at various levels. We begin
by discussing its implementation at grand strategic,
military strategic, operational, and tactical levels of
command. There are many other levels within each



Confrontation Analysis

of these four; however, discussion of these broadly

defined levels will present the general idea.
Then follows a discussion of what confrontation
strategy conducted on these four levels consists of.
To illustrate, we use two examples: the roadblockremoval in tables 2 and 3 found in chapter 2 and
General Deloittes strategy in chapter 6.

How Different Levels of Command Implement a

Confrontation Strategy
First, the grand strategic or political level is that at
which, in the play, Western public opinion was seen
to be putting pressure on Western governments to
stop ISRA from massacring people. A confrontation
strategy at this level is primarily the responsibility of
politicians. They are dealing with the public, with each
other, and with foreign governments. As in all
confrontations, they are trying to orchestrate the
emotions and arguments provoked by various
dilemmas to get everyone to play their tune.
Some of the cards politicians can play are military;
therefore, the militarys role is to advise them on the
practicability and likely consequences of playing one
or another military card. The military also needs to
understand the grand strategic confrontations that
politicians are involved in to understand the intent of
their superior and their superiors superior.
At the military strategic level, senior military staff
maintain and deploy a nations total defense forces in
pursuit of national political objectives. Having received
directions from the grand strategic level, this level must
issue strategic directives to operational-level

Chapter 7


commanders. In our terms, these directives empower

a commander, in the case of an OOTW, to implement
a confrontation strategy against certain other players
in pursuit of certain objectives. They also give him
certain types of cards he can play; therefore, it is
necessary for staff at the grand strategic level to have
a broad understanding of the confrontation strategy
required of their OOTW commanders to pursue. In
addition, they are involved in various, hopefully
cooperative, confrontations at their own level (e.g.,
among themselves or with the military staff of other
countries with whom they are required to act in coalition).
The next level, the operational level, is the level at
which General Deloitte is operating in the play. Having
received his strategic directive from the British national
defense staff, his task is to play his cards in such a
way as to achieve his mission objectives. To do so,
he must confront various local parties and try to get
them to comply with what he wants. In the play he is
planning to do this and, at one point, actually leaves
the room to do it. In this, he is implementing a
confrontation strategy.
At the operational level General Deloitte, like the staff
at grand strategic level, will conduct various other
confrontations, hopefully cooperative, such as
horizontal ones with other units in a joint or combined
force and internal ones with his own staff. In the play
we see him in a sequence of such confrontations with
his chief of staff and Mark Bright. (These personal
confrontations are not analyzed, just acted out.)
Finally, the generals confrontation strategy needs to
be analyzed and broken down into strategies that can
be carried out by units reporting to him at the tactical


Confrontation Analysis

level. The significant thing about the tactical level is,

in general, that this is where things, physical things,
actually get done.
What things? In the case of battle-fighting, tacticallevel units carry out and support destructive military
operations. In the case of an OOTW, destructive
military operations are not necessarily required.
When they are, their primary function often is to send
a message rather than to cause physical destruction.
Non-destructive operations, such as assisting
refugees and helping with disaster relief, may be
required. Even then, the ultimate objective is to get
others to carry on such operations, perhaps after
receiving initial or interim help, rather than for the
military to do so indefinitely.
Generally it is true in an OOTW that the objective is
not so much to do physical things as to get others,
those on the spot who will remain after we have gone,
to do or refrain from doing things. (Operations such
as disaster relief may seem to be an exception. Here
there are certainly physical things to be accomplished.
Yet even here the need to obtain cooperation with
other agencies and local bodies is often paramount.)
In a sense the above is true of all military operations.
Clausewitz himself says that after having render [-ed
the enemy] incapable of further resistance we compel
[him] to fulfill our will (1968, 1st edition 1832; p. 101).
In Clausewitzs traditional view, the military task, as
such, was to render him incapable of further
resistance. After that was done, politicians took over
to dictate to him our will. In an OOTW, by contrast,
the militarys task is to induce or compel others to fulfill

Chapter 7


our will, if possible without first rendering them

incapable of further resistance.
This is the objective and essential principle of a
confrontation strategy. In relation to what is done at
the tactical level, the point is that while we are
conducting a confrontation, we are not essentially
doing anything except communicating. Here
essentially means that we may be doing other things,
but all such other things will have communication as
at least one of their functions, and it is this function
that our confrontation strategy is concerned with.
In devolving a confrontation strategy to the doing, or
tactical level, we again must devolve, not a number of
physical tasks as such, but a number of cards to be
used in confronting others and getting them to comply
with our objectives. The essential difference,
nevertheless, between the operational and tactical
levels is that tactical level confrontations generally are
conducted at the local, grassroots level.
In devolving an operational level strategy, the essential
principle is as follows: To plan tactical-level
confrontations to change others parties foot-soldiers
in such a way that the message they send up to their
theater-level commanders reinforces the message we
are sending to those commanders at the operational
level. This is the essence of how to devolve a
confrontation strategy to the tactical level.


Confrontation Analysis

Linked Confrontations in a Confrontational

We see that an OOTW campaign is not a single
confrontation, but a number of confrontations at
different levels linked to each other. Following is a list
of some of those confrontations:
An operational commanders mission is defined
for him at the military strategic level after
politicians in a confrontation at the grand strategic
level have decided to play certain military cards. It
is the political confrontation at the grand strategic
level that defines his superiors intent and his
superiors superiors intent.
The operational commander is then involved in
various (hopefully cooperative) horizontal
confrontations with other units in a combined or
joint force; he must resolve these to properly
deploy his military assets.
His main confrontation is with other parties that
must be brought into compliance with his
mission objectives.
Finally, the operational commander must devolve
his main confrontation strategy to the tactical
units he commands. These then face numerous
local confrontations at the tactical level.
We are now close to what a confrontation strategy is.
But first we need to say how a theater commander
might, in a simple, step-by-step process, form and
implement a confrontation strategy for his main
confrontation, that in which he must get other parties
to comply with his mission objectives. Links with other

Chapter 7


confrontations will be discussed where relevant to this

description of a confrontation strategy in action. After
describing a confrontation strategy in action, we can
say in general what a confrontation strategy is.
Following is a step-by-step description.

A Confrontation Strategy in Action, Step by Step

In step 1, a commander builds a model of his
confrontation and identifies within it the dilemmas
faced by all parties. These include his own and others
weak and strong points (i.e., the points at which he
and others are open to being changed).
In the play, Mark Bright helps General Deloitte and
his chief of staff build a model, which is essentially
built and owned by the general, with all substantive
assumptions belonging to the general, while Mark
provides expert advice and guidance on what a
confrontation model is. Mark might argue for or
question a certain assumption on the basis of his
knowledge of the theoretical question being answered
by it. He does this quite often in the play. After he is
sure the general knows the meaning of a question, he
always accepts the generals answer. After arriving at
a first model, Mark helps the general identify that he
faces a deterrence dilemma. He does not, as perhaps
he should, go on to check that he has identified all
dilemmas; instead, he immediately asks the general
to consider what to do about this one. He is taking
short cuts based on his knowledge of model building.
After all, the general asked him to do what he can in
an hour.
In the model shown in table 2, the commander might
identify himself as facing a trust dilemma and the ethnic


Confrontation Analysis

militia as facing a cooperation dilemma. On the other

hand, the commander does not have a threat dilemma.
Rather than allow the militia to keep roadblocks in
place, he would prefer to remove them forcibly. His
first preference would be that the militia should remove
the roadblocks themselves.
In step 2, after the commander has identified the
dilemmas in his model, he chooses a subset of
dilemmas he will attempt to eliminate and
simultaneously establishes the means by which he
will attempt such elimination. He must use his
judgment and knowledge of the situation to estimate
how much friction would have to be overcome in
bringing about various dilemma-eliminating changes.
General Deloitte decides to eliminate his deterrence
dilemma by introducing a new card, Blame the
government, and making it part of his fallback
position. Realizing that this will give him a threat
dilemma, he next decides to make the Moryan
president believe he will feel bound to play this card
if the president does not back a cease fire. He spends
some time discussing the feasibility of these changes
(i.e., how much friction he must overcome in
introducing this card and making the president believe
he will play it). Notice that the two dilemmas he is
eliminating are connected: the second is brought
about by the elimination of the first. Notice too that
there is a strong connection with the grand strategic
drama. Bright does not analyze it, he does not have
time, but the connection is as follows. The Moryan
president is trying to appeal over the generals head
to international opinion and Western governments
to get them to send directives to the general that will
change his preferences (i.e., make him prefer the

Chapter 7


Moryan position (M) in table 8 to the threatened future

t. The general will have to prefer it, the president
reckons, if he is directed to implement it.) In this way,
the president is trying to overcome a deterrence
dilemma, consisting in the fact that the general is
under no pressure to accept the presidents position.
The president is trying to overcome this dilemma by
making moves in the grand strategic confrontation
that will, he hopes, affect the present drama through
linkages between the two confrontations.
In table 2, the commander finds he has a problem.
The militia has signed the agreement but is continuing
to maintain roadblocks. When questioned, the militia
leadership denies that there are roadblocks or blames
them on local groups. Meanwhile the militia points
out that the commander himself is violating the
agreement by withholding aid and support. Shall the
commander give up on the agreement and move to
the threatened future t (i.e., forcibly remove all
roadblocks)? (This was IFOR policy after the Dayton
agreement.) Or shall he try to change the militias
preference for unofficially keeping roadblocks in
place? He decides to try the latter. He devolves his
strategy to the tactical level, directing local
commanders such as the one in table 3 to persuade
local militia to remove roadblocks. He tells the militia
leadership that this is what he is doing. He is, he says,
helping them to make their policy effective with local
groups so that aid and support can start to flow.
Simultaneously, he starts a campaign on local
television telling local militia groups what is the official
policy of their leadership. He reasons that if local
commanders, thus supported, succeed in their local
confrontations, then the leadership of the ethnic militia


Confrontation Analysis

will change its preferences and abandon its unofficial

policy of maintaining roadblocks. (Note that this is a
link that he sees between the two levels.) He will then
let them have the promised aid and support.
Step 3 is to implement the planned dilemmaeliminating measures. Note that by definition no one
can know, when entering a moment of truth, how they
will be changed by it. This applies less to the
commander, who is deliberately instigating a moment
of truth to cause specific dilemma-eliminating changes,
than to those he is interacting with, who are, in
comparison with him, unprepared for the pressures
they are about to face; nevertheless it applies in some
degree to both sides.
In the play, General Deloitte gets on the phone to
President Saldin and lets him know he had better come
to the cease-fire talks and back a cease fire, or he will
be publicly blamed for being responsible for the
massacre. This, he must see, will undermine his efforts
to enlist Western support. President Saldin now faces
a moment of truth.
The UN commander in table 2 sends directives to his
local commanders, describing his confrontation
strategy and how they can contribute to it, giving them
appropriate cards to play but leaving the details of
their individual strategies to their initiative as
commanders. If they have questions or suggestions,
such as additional cards they might want to have in
their hand, they discuss them with him. At the same
time he informs the militia leadership of what he is
doing, presenting it as a way of helping them to exert
control over local groups and holding out to them the
prospect of aid and support. Finally, he launches a

Chapter 7


television campaign while his local commanders

confront local groups manning roadblocks. In this way,
the militia leadership is brought to a moment of truth.
Step 4, following the moment of truth, at which
changes were meant to take place, is a return to step
1 to build a model representing the new, changed
situation. If there is now a satisfactory solution, with
all parties taking the same position and trusting each
other to carry it out, then the commander has achieved
his mission objectives (perhaps in a modified form). If
not, dilemmas remain, which the commander identifies
at this point. He then moves on to step 2, and so on.
In the play we saw how the commander, having
brought the Moryan government into compliance with
his objectives, models the new situation and works
out how to do the same to ISRA.
Suppose, in table 2, that the UN commanders
strategy succeeds in getting the militia to comply
with the removal of roadblocks. There are now no
more dilemmas. The problem is resolved within this
model. Other, more detailed models, or models
focusing on other issues, would no doubt reveal
difficulties. In addition, implementation of the
agreement may meet unforeseen problems. For
example, the commander may find that the aid and
support he promised to the militia, and which he has
held up by refusing to certify that the first stage of
pacification has been reached, is now held up for
other reasons. To get around this problem, he must
resolve a horizontal confrontation with the high
representative responsible for aid coordination.


Confrontation Analysis

Note: A confrontation between two parties is vertical

if they are related as superior and subordinate in the
same organization. This means that one party
generally has a card consisting of formally issuing
orders or directives to the other, although often this
does not settle the matter; if the other does not agree,
it may be able to ignore or re-interpret directives given
to it, or appeal against them. The distinction between
vertical and horizontal relations becomes fuzzy when,
as is often the case in peace operations, a commander
is responsible both to his national headquarters and
to an international organization. For the sake of our
example, we are assuming that the high representative
has no authority over the UN commander (as in the
case of the Dayton agreement, where the UN high
representative was responsible only for coordinating
civilian aspects of the agreement). Therefore,
negotiations between the UN commander and the high
representative in our example are horizontal. Neither
has authority over the other.

The Confrontation Strategy Itself

The above step-by-step description of the
implementation of a confrontation strategy is not yet a
description of the strategy itself, but it enables us to
give the following definition. A confrontation strategy
consists of a plan for cycling through the above four
steps until a resolution of the confrontation is reached;
therefore, it consists of a sequence of dilemmarevealing models of characters positions such that:
A plan for dilemma-elimination is associated with
each model except the last.

Chapter 7


The next model in the sequence is created by

implementing the plan on each model except
the last.
The last model has no dilemmas.
This is our suggested characterization of a
confrontation strategy. To understand it correctly,
several points must be kept in mind.
First, each plan for dilemma-elimination associated
with a model in the sequence is a plan for bringing
about a moment of truth, at which changes may take
place. We have said that on entering a moment of
truth no one can tell exactly how it may change them.
This, as well as the general fact that the future tends
to hold surprises, makes it an essential characteristic
of a confrontation strategy that it should be looked at
anew, and reformulated if necessary, at each stage,
after each moment of truth.
Even though it is constantly reformulated, it is
nevertheless desirable for a whole strategy to exist,
so that assets can be mobilized and relevant
information collected to fulfill it.
Second, it is important to realize that a plan for
dilemma-elimination associated with a particular model
in the sequence may be linked to other, related
confrontations. We have looked at some examples of
such linkages between the particular confrontation a
commander is involved in and other confrontations.
For example, the confrontation between General
Deloitte and President Saldin was linked to the grand
strategic confrontation between the West, the Moryan
government, ISRA, and other countries (e.g., the Arab


Confrontation Analysis

nations, France, the bordering state of Pelugya). Part

of this linkage consisted in the fact that since General
Deloittes job was to fulfill his mission, his preferences
were largely linked to the policy decisions of the West.
If Western politicians wanted him to side militarily with
the Moryan government, that likely would become his
preference. The result was that to eliminate a dilemma
in his confrontation with the general, President Saldin
could try to alter the latters preferences by means of
moves in the grand strategic confrontation he was
conducting with the West.
The confrontation in table 2 between the UN
commander and the ethnic militia leadership was
similarly linked to the confrontations between local
commanders and local militia groups; that is,
decisions by a superior could decide the preferences
of a subordinate. However, interestingly, the opposite
was also true in the case of the ethnic militia. The
commanders plan for eliminating a trust dilemma
depended on lower-level militia attitudes influencing
the preferences of their leadership.
After apparently resolving his confrontation with the
ethnic militia, we saw the UN commander finding
himself unable to implement the agreement reached
because of problems with a horizontally related
partner, the high representative responsible for
coordinating aid. Technically this meant the
commander had to resolve a horizontal confrontation
to eliminate a cooperation dilemma.

Chapter 7


Applying the Principles of War

to Confrontations
It is clear that a confrontation strategy is different from
a strategy for warfighting, yet some of the same
principles apply; therefore, it may be useful to ask
which generally known principles of war continue to
apply to confrontations. This may help to bring out
what is different about OOTW. We are not attempting
any more than this; we do not aim to establish a list of
principles of OOTW.

Seven Principles of War

Alberts and Hayes (1995, pp. 28-37) ask how far the
principles of war apply to peace operations. They
survey seven principles of war and conclude that
simplicity and unity of purpose (rather than of
command) remain important. Unity of command, they
point out, is often unachievable when there is a need
to coordinate strategy between different national
contingents; it has to be replaced by the attempt to
achieve unity of purpose. Achieving this also helps to
focus on the objective. Consequently they suggest that
unity of purpose can replace this principle also.
Thus, Alberts and Hayes endorse simplicity as a
principle for peace operations and subsume the two
principles, unity of command and focus on the
objective, under the one principle, unity of purpose.
They suggest adding consensus planning, adaptive
control, and transparency of operations as principles
to be followed in peace operations. Four other
principles that apply in warfighting, taking the offensive,
concentration of superior force, taking the enemy by


Confrontation Analysis

surprise, and maintaining security, are, they consider,

less applicable and may even do harm.
This is certainly right if we look at the physical
operations involved, as we do when considering
warfighting. We have said, for example, that a
physical operation such as a bombardment or a
troop movement whose primary purpose is to send
a message cannot sensibly be kept secret. Thus
the principles of surprise and security, applied in
ways that would make sense if the primary purpose
were destruction of the enemy, can do harm if
applied in OOTW.

The Contrast Between Warfighting

and Confrontation
The contrast involved here is important; it perhaps
needs re-emphasizing. If we are bombarding the
enemy to force negotiation, our bombardment is
actually a message to the effect: Look what will
happen to you if you dont come to the bargaining
table. Its physical effects in actually destroying the
enemys assets (as distinct from communicating our
determination to do so) are a disadvantage to us as
well as to the enemy, because the more assets we
have already destroyed, the less remain to be
destroyed, and the less significant is our threat to
continue destroying them.
Counter-insurgency forces, for example, by destroying
the homes and families of guerrilla-fighters, often
create large numbers of homeless, bereaved fighters
with nothing left to lose and much to revenge.

Chapter 7


Consequently, to send the message, look what will

happen to you, the ideal is to combine a maximum of
shock and awe with a minimum of actual destruction.
Keeping secret the destruction you have wrought is
counter-productive. By contrast, if the intention is not
to send a message but to actually destroy the enemys
fighting potential, then the more secure he feels and
the less he knows about what is happening to him the
better, because his ignorance and contentment reduce
the effectiveness of his countermeasures and
maximize the damage done to him.
On the other hand it may be a sensible strategy, as
we have seen, to interrupt negotiations to weaken
anothers bargaining position by physically taking
certain cards out of the other partys hand. To
complicate matters, such physical operations may
be combined with message-sending. For example,
operations such as interdicting terrorists
communications or destroying their arms dumps may
reduce their ability to harm us, and thus either take
cards from their hand or reduce the effectiveness of
certain cards they might play. Such operations
obviously require secrecy and surprise at the tactical
level, but if carried out while we are conducting a
confrontation with terrorist leaders, they may serve
two functions: to send a message (This is what you
can expect if you dont concede) as well as to diminish
the effectiveness of terrorist action. The message may
even be, You can expect your effectiveness to keep
diminishing as long as you hold out.
An operation that has dual functions, messagesending combined with physical effects, may need a
balance of secrecy and openness. It remains true that
a message as such needs to be clearly received and


Confrontation Analysis

correctly interpreted if it is to do its job. To destroy

terrorists effectiveness will not work well as a message
if the terrorists think they have suffered less loss of
effectiveness than they actually have, or mistakenly
believe that we mean our strikes against them to be
our last. Yet these mistaken beliefs actually would
benefit us if our actions were not meant to send a
message, but were carried out purely to diminish the
terrorists physical effectiveness.
The contrast between something done for its actual
effect and the same or a similar thing done to send a
message is striking. We have said that a confrontation
strategy (the multilevel, coordinated object described
in the last section) is not essentially a matter of
conducting physical operations, but of sending
messages. It is not surprising that we get somewhat
different results when we look at how the AlbertsHayes principles of war apply to message-sending
rather than when they are applied to physical activities
as such.

Applying the Principles to a Confrontation Strategy

Consider now the seven principles in turn.

Simplicity is of utmost importance, essentially for the

reasons given by Alberts and Hayes, but we can add
some more. We have seen that an operational
commanders confrontation strategy needs to be
understood and implemented in a coordinated way
by tactical echelons below him and other task-force
units horizontally related to him. He also needs to be
able to explain it clearly to military strategic and grand
strategic levels above him to enlist their support in
pressuring parties from above or expanding the cards

Chapter 7


available to him, when necessary. This is the

confrontation-strategy need for simplicity. Simplicity
is also necessary because the common reference
frame through which the commander communicates
his position to other parties must be simple; therefore,
his confrontation strategy can be simply expressed in
terms of it. The commander in table 2, for example,
might explain his strategy by presenting and explaining
tables 2 and 3.

Unity of purpose, including, as Alberts and Hayes

suggest, focus on the objective, also magnifies the
effectiveness of a confrontation strategy. Suppose a
commander is pressuring another party to take a
certain action, say to cease fire, withdraw armaments,
or permit the passage of refugees, and to this end is
using certain implied threats and promises. We have
seen that he is more likely to succeed if the other
partys grassroots organization, superiors, and allies
are receiving versions of the same message, and so
are inclined to tell it to make the concession. Such a
unanimous chorus is achieved only by getting unity of
purpose behind a message dictated by an agreed
confrontation strategy. We have also seen that
obtaining agreement between different coalition
partners on a common confrontation strategy is itself
a task for a (separate, hopefully cooperative)
confrontation strategy.
Taking the offensive is shown by Alberts and Hayes
to be an inappropriate principle for peace operations,
if we think in terms of a physical offensive; however,
as applied to a confrontation strategy, we might
reinterpret it to mean taking the initiative in forcing a
moment of truth so that a certain set of dilemma
eliminations, planned by us in advance, can occur.


Confrontation Analysis

This is to take and hold the initiative in terms of the

argument. It is analogous to the battle-fighting concept
of forcing the enemy to respond to us, rather than vice
versa. To avoid confusion, it might be better to avoid
metaphor and call it taking the initiative rather than
taking the offensive.
Is concentration of superior force another principle we
can usefully reinterpret in application to a confrontation
strategy? Superior force might be taken to mean an
array of superior arguments for use at a preplanned
moment of truth, provided we understand that the
arguments we have in mind need not be very
intellectual, but may include points such as Well bomb
you to bits if you dont. Here argument may be the
wrong word. Force may also be misleading.
Sanctions is perhaps the right term, because it may
refer to positive enticements such as offers of aid as
well as negative ones. This suggests that
concentration of superior sanctions might be the right
name for this principle.
Concentration of sanctions may be unnecessary
because of what appears to be an essential difference
between confrontations and battle-fighting. In a battle,
assets used in one place cannot at the same time be
used in another; therefore, they must be concentrated
before an assault. By contrast, negative assets used
in a confrontation, that is, the cards you can play to
out-escalate another, are not necessarily used. The
hope is that you will merely have to threaten their use.
The same negative assets can therefore be used in
different places at once. Following are two examples:
All the commanders confronting local militia in
our roadblocks example can simultaneously

Chapter 7


threaten to call down air strikes. They will not

need that many planes.
The British Empire, while it lasted, was held
down with a small number of troops. Each
potential rebel feared that those troops would be
used against him.
Positive assets (sanctions that are promises rather
than threats) are different. They will be used up in a
successful operation because using them up is, by
definition, the objective. Their quantity must be
sufficient to cover the different points where they
are used.
In sum, it seems that assemblage of superior sanctions
may be the principle that should replace concentration
of superior force.

Surprise is another principle inapplicable to OOTW

when we consider only physical operations. Given the
need to build trust and verify agreed physical
conditions (such as separation of forces or
disarmament), peace forces should not create
uncertainty by behaving in unexpected ways (Alberts
and Hayes, 1995, p. 30). Indeed, this follows from our
assumption that actions in a confrontation are seen
primarily as communications. Our argument is as
follows. One of the characteristics of a surprise action
is that it is uninterpreted. You wonder what it means.
That defeats the purpose of an action that is supposed
to send a message. The whole point of that is for you
to know what it means. Prior notice tells other parties
what an action means.
The principle of surprise, however, may be applicable
to the content (i.e., meaning) of the messages you


Confrontation Analysis

send in a confrontation, as distinct from the form (the

action itself). Producing a surprising new argument,
offer, or threat, as General Deloitte in the play does
in threatening to blame the president for the
massacre, makes it hard for the other to produce a
convincing defense, because they have not been
given time. Lawyers in television courtroom dramas
win this way. In the other direction, a surprising
concession or compliment, even if insignificant, may
cause favorable changes in others perceptions of
your motives just because it is surprising. A surprise
draws attention to itself. The principle of surprise,
understood in this way, can be used alongside taking
the initiative and assembling superior sanctions. They
are mutually reinforcing methods of obtaining other
parties willing compliance.
Finally, Alberts and Hayes cast doubt on the principle
of security , pointing out how over-emphasis on
physical security, e.g., ensuring minimal risk to troops,
may reduce the effectiveness of peace operations.
This follows from asking, What message does this
action send? Strict security precautions say, I dont
trust you. This is not compatible with Trust me. On
the other hand, it is quite compatible with the message,
Im ready to smash you, which may, in other
circumstances, be the intended message. In relation
to physical security, all that confrontation analysis can
say in general is what it says in relation to other
physical activities: Pay attention to the message you
want to convey.
On the other hand, surprise requires security, and we
have said that surprise may be advantageous in itself
and help us to take the initiative and assemble superior
sanctions; however, this applies to the content of

Chapter 7


messages, not to physical actions. To ensure surprise

in our confrontation strategy, and also to preserve
simplicity and avoid misunderstandings, it is advisable
to preserve security in the process of confrontationstrategy formation; however, in implementing a
confrontation strategy, disclosure of our strategy is
usually the best tactic because it reinforces the
message being sent and helps to avoid
misinterpretation. Similar reasons are given by Alberts
and Hayes for their principle of transparency of
operations. They are the reasons why, in our example,
the UN commander decides to inform the ethnic militia
leadership of his devolved strategy of confronting local
militia roadblocks.
This seems to leave us, in this area, with a rather
awkward hybrid slogan: Security of planning plus
transparency of operations.
Summing up, the principles of war apply rather badly
to OOTW considered as physical activities; however,
many of them seem to apply better to the formation
and implementation of a confrontation strategy, which
is primarily a matter of sending messages.

Summary of Chapter 7
A confrontation strategy for a particular confrontation
consists of a plan for sequentially eliminating
dilemmas, each in a specific way, so that the end result
is full compliance on the part of all parties with our
mission objectives. A strategy for a peace campaign
as a whole is a plan for winning in this manner a linked
set of confrontations.


Confrontation Analysis

A strategy needs to be adaptive to changing

circumstances and new information, such as the fact
that a particular confrontation is not resolved in the
manner expected. At the same time it needs to be
sufficiently robust to give continuing guidance
throughout the inevitable fluctuations of expectations
that occur in the course of a campaign.
For these reasons, a strategy must have a clearly
stated objective; be simple; and be carried out with
unity of purpose. As has been pointed out, other
principles of war, such as surprise, concentration of
force, taking the offensive, and security do not apply
to peace operations in the same way as they do to
warfighting; however, applying superior force and
taking the initiative are applicable to confrontation
strategy at the level of persuasion, as distinct from
physical operations.
A strategy needs to be understood and implemented
at various vertically linked levels of command and by
various horizontally linked components and coalition
forces. We must understand the manner in which
different confrontations are linked to formulate a
confrontation strategy. A confrontation is linked to
others that its conditions and objectives affect or are
affected by. It is linked to higher-level confrontations
that set the scene for it and to lower-level
confrontations that its Implementation phase sets the
scene for; also, a confrontation is linked to the more
detailed confrontations that take place in its Resolution
and Conflict phases.
Confrontations are linked together strategically if the
positions and the sequence of dilemma-eliminations
adopted in one confrontation are linked to those

Chapter 7


adopted in others. In the removing-roadblocks

example, a confrontation strategy worked out by the
theater commander could be devolved to the level of
the local commander by linking their confrontations.
Similarly, the commanders confrontation with noncompliant parties at theater level should be linked to
those of horizontally related characters such as other
component and coalition forces. Achieving
confrontation strategies that are coordinated between
horizontally and vertically linked partners may require
consensus planning and negotiation of strategies.
Above all, it requires that the commanders strategy
be simple and his objectives clear and agreed.

Chapter 8
Analysis of the
Bombardment of
Sarajevo, February 1994

n February 5, 1994, a bomb exploded in

Sarajevos market square. It killed 69 people and
left 200 wounded. Blame naturally fell on the Bosnian
Serbs, who at the time were bombarding the city with
mortars placed on the surrounding hills. They at once
denied responsibility and blamed the Muslim
defenders of the city. Nevertheless, there was an
international outcry against the Serbs. As a result, the
British commander of the United Nations Protection
Force (UNPROFOR) in Bosnia, LTG Sir Michael Rose,
found himself tasked with getting the Serbs to cease
fire and withdraw their weapons in order to fend off
NATO air strikes on their positions.
In this chapter we examine this crisis in some detail,
going through the most complex confrontational
analysis of a real-world Operation Other Than War
(OOTW) that has been done so far. As we do so, we
will discuss some of the problems of real-world
confrontation analysis.



Confrontation Analysis

Sources for the Analysis

Research for the analysis of Sarajevo, February 1994,
was carried out by the author in early 1997 under a
contract issued by the UK Defense Evaluation and
Research Agency (DERA) for a report on the use of
drama theory in operational analysis (DERA, 1997). It
is reported here by permission of DERA.
The research included interviews with a number of
British officers who had served in Bosnia; however,
they were not interviewed for their involvement in the
particular confrontation analyzed, the crisis over the
Bosnian Serb bombardment of Sarajevo in February
1994. All factual assumptions made in the analysis
are drawn from published sources. If they are
inaccurate, the fault lies either with those sources or
with our misinterpretation of them, and we can only
hope that those in the know, in particular General Rose
and his staff, will make allowances accordingly.
Even if we had been able to interview actual
participants, they would not have been the best source
for assumptions to go into our analysis. Memories are
biased, rationalized reconstructions of what was felt
and thought at the time. We tend to think that what
did not happen, could not have happened. We do not
think so at the time. The best source of assumptions
for confrontation analysis is the judgments of those
involved while the confrontation is going on. This is
how assumptions are made in the fictitious case
described in the Frontline play in chapter 6.
Incidentally, the reader will notice that the play,
although fictitious, has a plot that is based on the crisis
in Sarajevo in February 1994. The play was written by

Chapter 8


taking that real-life scenario, simplifying it, equipping

it with imaginary characters, transferring it to a nonexistent North African location, and imagining that it
was dealt with by a commander who decides to use
confrontation analysis.

Conclusion of the DERA Report

The conclusion to keep in mind and examine critically
as you read this chapter and later ones is as follows:
If the UNPROFOR commander in Sarajevo had used
confrontation analysis in his efforts to get the two
warring parties to cease fire, he might have achieved
his mission objectives more quickly and satisfactorily.
In the DERA report, the conclusion is spelled out as
The first stage, bringing the Bosnian government
into apparent compliance, might have been
achieved in a way that was planned and
intended beforehand, without depending on
emotional, spontaneous, unplanned reactions of
the commander. Emotions might have been
used, but in a rationally planned way.
The second stage, bringing the Bosnian Serbs
into apparent compliance, might have been
achieved sooner by the UNPROFOR
commander on the operational level instead of at
the last moment, after many concessions had
been made, as a result of developments on the
grand strategic level.
Even if the result had differed little from the
actual one, it could have been achieved by the
commander in a more purposeful way because


Confrontation Analysis

from the beginning he would have been able to

seize and maintain the initiative.

Putting Pressure on the

Bosnian Government
When the bomb exploded in a square crowded with
shoppers, the international outcry brought to a head
a growing demand for intervention to stop the Serb
bombardment of Sarajevo. The UNPROFOR
commander actually was engaged at the time in
negotiations to this end, tackling the problem that
while the Serbs were predominant in heavy
weapons, they were weaker in infantry than the
(mainly Muslim) Bosnian government forces
defending the city, who were ready to take the
offensive if Serb actions were halted.
At this time, the UNPROFOR commander conceived
his mission as primarily humanitarian (delivery of
aid, prisoner exchanges). After the bomb fell, he went
to Belgrade, met with the overall UNPROFOR
command and the UN special attach, and came
back with a mission to get the Serbs and the Bosnian
government to agree to a cease-fire and withdrawal
of heavy weapons.
Simultaneously, at the grand strategic level, the U.S.
and French governments began pressing for a NATO
ultimatum to the Bosnian Serbs: Cease fire and
withdraw heavy weapons or we launch strategic air
strikes against you. Although it supported his efforts
to get Serb agreement, this was of serious concern to
the UNPROFOR commander, who feared that NATO
air strikes would make it impossible to continue his

Chapter 8


humanitarian mission and would endanger the lives

of UN and non-governmental organization personnel.
His reasoning followed this path: If my mission is
humanitarian, let me do it. If you want to change my
mission to fighting the Serbs, I must withdraw UN
personnel and abandon the humanitarian mission.
The commander proceeded urgently with his new
mission in the hope of forestalling air strikes. He had
already laid the groundwork. In previous discussions
he had obtained the agreement of commanders of
Serbian and government forces to four points: a
cease-fire; withdrawal of Serbian heavy weapons to
at least 20 kilometers from Sarajevo if not placed under
UN control; the interposition of UN troops; and daily
meetings to oversee implementation of these
agreements. When the crisis broke, however, Bosnian
President Alija Izetbegovic decided he would now
make no agreement with the Serbs. Rather, his policy
would be to take advantage of the outcry to persuade
the international community to take action on the
Bosnian governments behalf, specifically by lifting the
arms embargo against it and taking military action such
as air strikes against the Serbs.
The result was that the commander discovered that
the president, who was at that moment being
interviewed by CNN and declaring the need for air
strikes against the Serbs, had told his commanders
not to attend the UNPROFOR-arranged meeting.
Angrily the commander let it be known that if the
Bosnian government did not attend the meeting, he
would publicly blame Muslims for the market-square


Confrontation Analysis

incident and for blocking a cease fire. Table 11 is our

attempt at a card-table model of the commanders
moment of truth at this juncture.
In table 11 the symbol ~ indicates that a character
takes no position as to whether a particular card should
be played or not; that is, its position is that the card
may or may not be played. The Bosnian Serbs, for
example, took the position (column BS) that
UNPROFOR should not blame them for the marketsquare bomb nor call air strikes against them, while
they themselves should not cease fire or withdraw
weapons; in return they would not retaliate against
UN personnel. They took no position as to whether
UNPROFOR should blame the Muslims or the Bosnian
government should cease fire; obviously they would
have preferred both these cards to be played, but could
hardly demand or expect them, given their own
proposed actions. The numbers indicate preference
rankings as before, with the most preferred future given
the number 1, the next most preferred the number 2,
and so on. In assigning preference rankings to columns
containing entries with ~, the most probable decision
is assumed. For example, it is assumed in this table
that if the Bosnian Serb position, BS, were accepted,
then the players expectation in regard to the cards
left unspecified (Blame Muslims and the Bosnian
governments Cease fire card) would be that neither
card would be played: UNPROFOR would not blame
the Muslims and the Bosnian government would not
cease fire. This is why the column BS is assigned the
same preference ranking for each player as column
d, even though it is compatible with either t or d.

Chapter 8


Table 11. Bosnian government refuses to discuss a cease-fire.


Confrontation Analysis

The following assumptions are made in the model:

The UNPROFOR position (column U) was that
both sides should cease fire while U
(UNPROFOR) remained impartial, not blaming
or attacking anyone. Note that it was important
for U that BG (the Bosnian government) as well
as BS (the Bosnian Serbs) should cease fire. BG
had more infantry, although fewer heavy
weapons, than BS. Also, it often initiated
incidents, even though it might lose militarily, to
get world opinion on its side.
The commander could assume that the BS
position was as shown (column BS), even
though BS had agreed to Us four points. This is
because BS had agreed unwillingly, arguing that
as BG forces outnumbered theirs, they needed
to use artillery to respond to BG-initiated
incidents. When BG rejected the four points, it
could be assumed that BS would reject them
also in favor of its previous position.
The BG position was that they, the victims,
should not be asked to cease fire. U should take
their side, blaming BS and calling air strikes
against them. BS could do as they wished, (i.e.,
BG took no position as to which BS cards should
be played).
The assumption behind the fourth column is
that, by angrily interrupting the presidents
CNN interview and directly threatening him, the
commander made the threatened future one in
which he would publicly blame the Muslims
and the bombardment would continue. This, of

Chapter 8


course, was very much what the Serbs

wanted. The threatened future put no pressure
on BS; however, it did pressure BG to accept
Us position.
The fifth column states that, if present policies
and actions continued unchanged, there would
be no cease fire and U would take no action.

Other Parties in the Background

There were at least two other important parties in the
background of the drama we are analyzing: the
Serbian government of Slobodan Milosevic and the
Croatian government of Franjo Tudjman. At this point
both were in transition from one policy to another. The
Serbian government was attempting to get
international sanctions against it lifted by distancing
itself from the Bosnian Serbs, who had advanced so
far with its backing. The Croats were engaged in U.S.sponsored negotiations to restore their alliance with
the Bosnian government against the Bosnian Serbs,
having in the most recent phase of the conflict been
taking land and expelling Muslims in unspoken alliance
with the Serbs. Alliances were reforming against the
Bosnian Serbs. This, we assume, increased their
feelings of persecution and discrimination while making
them more anxious to take Sarajevo while they could.
At the same time, it meant that the Bosnian and
Croatian governments played no active role in the
confrontation we are analyzing.
Note that we are following and will continue to follow
an important rule of realistic modeling, as stated in
the play in chapter 6 by Mark Bright while advising the
general. Mr. Bright points out that factors outside the


Confrontation Analysis

model that influence things in it must be taken into

account in making assumptions. They should not be
ignored just because they are not explicitly named in
the model.

Analysis of Table 11
What dilemmas face the characters in this model?
Take the UN commander first.
U has no deterrence dilemma against BG. BG prefers
column U to t ( t would undermine international
sympathy for BG); therefore Us threat, if credible,
puts pressure on BG to accept column U. The result
is that the UN commanders threat concerned
Bosnian President Izetbegovic; he broke off his
interview with CNN.
U does have a deterrence dilemma against BS, who
greatly prefers t (with the West blaming the Muslims
while their bombardment continues) to U. The result
is that the Serbs were under no pressure, and would
have been delighted if Izetbegovic had refused to
attend the meeting.
U has a threat dilemma; it prefers not to blame the
Muslims for the market-square bomb. That would
mean giving up impartiality and possibly going against
the grand strategic policy of NATO and the UN. The
result is that the UN commanders anger, without
which the Bosnian president might have dismissed
his threat. This anger showed either or both of the
That Us preferences were changing in favor of
carrying out the threat (preference change)

Chapter 8


That U would carry it out regardless of whether

or not it preferred to do so (irrationality).
U has an inducement dilemma in that t is arguably
worse for U than BS, a fact that puts pressure on U to
accept BS. The reaction was anger and defiance on
the part of the UN commander as the BG position,
with accompanying need to threaten t, puts pressure
on him to abandon his objective and accept BS.
U has no positioning or cooperation dilemmas. It
prefers its own position to the positions of the other
characters and to any future it could reach from its
own position.
U has a dilemma of trust in that both BS and BG
would prefer unilaterally to defect from Us position
by breaking the cease-fire. It is even preferable for
both if both defect: preferable for BS because it wants
to keep up the pressure on Sarajevo, preferable for
BG because it wants to keep up pressure on the
international community to intervene on its behalf.
There is no immediate reaction to this dilemma. But
note what happens next.
BG gives in to Us pressure and accepts Us position.
BG and BS army commanders then attend a meeting
with U. In this meeting all agree with Us position;
however, even at the time U does not give much
credence to this agreement. U does not decide to give
up. Instead, U decides to monitor whether the
agreement is observed while continuing, with positive
emotion, to stress its desirability to both sides in an
attempt to persuade them to keep it. All the signs are
that U feels it has a trust dilemma.


Confrontation Analysis

Other Parties Dilemmas

We have looked at Us dilemmas. What about other
characters dilemmas?
All the dilemmas facing the parties are set out in table
12. It is striking that the Serbs face no dilemmas. This
is partly because their own position is what they want
to see happen and is such that they do not need to
trust anyone. They have no cooperation, trust, or
positioning dilemmas. Also, the tug-of-war between U
and BG has made the threatened future one they like,
although the other two do not. They therefore have
no inducement, threat, or deterrence dilemmas. They
are sitting pretty.

Table 12. Dilemmas facing characters in table 11.

The Bosnian government has three dilemmas:

Deterrence: The threatened future, t, pressures
no one to accept BGs position. Both U and BS
prefer t to BG. The reaction was abandonment
of position BG.
Inducement: BG prefers U to t. The reaction was
acceptance of position U.

Chapter 8


Trust: Even if U were to agree to BGs position,

BG might doubt whether they would carry it out,
given the UN commanders strongly expressed
preference for remaining impartial and not taking
sides. The reaction was possibly greater
willingness to abandon position BG.

How UNPROFOR Got AngryThe Model That

Preceded Table 11
It seems probable that the card, Blame Muslims, in
table 11 was thought of by the UNPROFOR
commander as something he might actually do only
when he realized the position the Bosnian government
was taking. To see this, we draw up the card-table in
table 13 showing the moment of truth before he thought
up this card.
Both Bosnian sides here prefer t to Us position (BG
does so because t places pressure on the international
community to intervene on BGs side); therefore, U
now faces a strong deterrence dilemma (i.e., it is
pressuring no one to accept its position). Us reaction
is anger and demonization of BG, motivating U to think
up the card, Blame Muslims, and make it credible as
part of the threatened future t shown in table 11.
Note that if we look at the analysis formally, it might
seem that U might equally well have demonized the
Serbs. However, the Serbs had at least agreed to talk.
The Muslim-dominated government was refusing to do
so. Also, the Serbs already had been so demonized by
the media and Western public opinion that U would need
little rationalization, and hence little demonization, to
justify turning against them.


Confrontation Analysis

Table 13. What made UNPROFOR

angry? The situation preceding table 11.

This, then, is an example of emotion and rationalization

leading to a change in the set of cards assigned to
characters in a frame.
In table 14 we set out all the dilemmas in table 13.
U has a deterrence dilemma (as said) and a trust
dilemma (for the same reasons as in table 11).
BS has no dilemmas.

Chapter 8


Table 14. The dilemmas facing characters in table 13.

BG has a deterrence dilemma (no one is

pressured to accept its position) and a trust
dilemma (U would not want to implement BGs
position, if it were agreed).
BGs dilemmas obviously gave it serious problems in this
table; however, it had a strategy to overcome them,
consisting of a strong appeal to the international
community, which had it in its power to change the mission
assigned to U and thereby change Us preferences. If it
could get U on its side, this would be enough to achieve
its position, as its position did not specify any particular
actions on the part of the Serbs (i.e., any Serb actions
would be compatible with its position).


Confrontation Analysis

After the Bosnian

Government Was Brought into
Apparent Compliance
A commanders mission objective, expressed in
terms of confrontation analysis, is to bring others
into compliance with his position. Now at this point
in our analysis, it might seem that BG, at least, has
been brought into compliance.
A commander must, however, distinguish between
apparent compliance (overt acceptance of his position)
and actual compliance (in addition, the intention to
implement that position). He must be aware that a
character facing a cooperation dilemma may solve it
by deceit (i.e., accepting a position while intending to
act in non-compliance with it).
In this case the estimate is that BG has been brought
into apparent compliance only, because if a ceasefire were implemented, they would prefer to break it.
Although the UNPROFOR commander has achieved
the bringing of BG into apparent compliance, BG now
has a cooperation problem and the commander a trust
problem in that BG will want to defect from position U.

Problems in Determining a Characters Position

Has BS also been brought into apparent compliance?
A superficial reading might suggest so. We have said
that all parties, including BS, have at this point formally
accepted U.
This would be a mistake, illustrating the fact that
determining a partys position by the rule, It is what
it says it is, may be harder in practice than it seems

Chapter 8


in theory. A real-life character has many

subcharacters saying different, contradictory things.
Even the same subcharacter may contradict itself.
What do we do then?
Judgment is necessary, but a general rule in
interpreting others messages is, Pay attention to their
emotions. One thing we know is that a character
whose emotional signals and rationalizations are
predominantly negative toward others, angry, defiant,
and so forth, must be taking a position opposed to
theirs. We know this because such emotions are not
appropriate to solving cooperation or trust dilemmas,
which are the only ones left between parties that share
a common position.
If a negatively emoting party formally states that it
shares your position, it is contradicting itself by its
negative rationalizations. If these are as open and
public as its formal statement of agreement, if they
come across more strongly than the formal statement,
and if they state a clear alternative position, they should
be read as stating the partys actual declared position.
In this light, consider the present case. At this point,
the signals being sent by BS and BG are quite different.
BG is showing signs (i.e., fear of the threatened future
and a conciliatory attitude) of the impact of the
pressures needed to make them comply. BS is
showing opposite signs. The NATO ultimatum to the
Serbs, Cease fire and withdraw or we bomb you,
was issued on the day that agreement apparently was
reached. Instead of reacting with fear, depression, or
conciliation, BS reacted with defiance, threatening to
shoot down 40 NATO planes in the first wave of attack.
More seriously, they threatened retaliation against UN


Confrontation Analysis

personnel. This, together with their continuing strong

arguments for column BS, meant that this was their
position. They were not even in apparent compliance
with U.
One interpretation, possibly that of the commander,
is that the NATO ultimatum itself, by infuriating BS,
had made them non-compliant. It is equally possible
that BGs compliance made the BS non-compliance
obvious. While BG was defying U, BS may have
thought it had nothing to lose by seeming conciliatory.
In either case, Serbian defiance as now revealed
means that the correct model of the moment of truth
at this point is that in our next model, table 15. Here
BS has de facto not agreed to Us position, even
though formally it has done so. Only BG has agreed.
In addition, the threatened future is that U will blame
the Serbs and call air strikes against them. BS then
will retaliate against UN personnel.

Modeling the Grand Strategic Drama

Note that the threatened future in table 15 comes from
the NATO ultimatum rather than from the words of the
UNPROFOR commander, who was strongly opposed
to air strikes because he feared retaliation.
The NATO ultimatum emerged from a grand strategic
drama going on in parallel to the operational drama
represented in table 15. We attempt to model this
higher-level drama in table 16, before discussing how
it is linked to table 15. Again, our assumptions about
the grand strategic model use published sources only,
and readers who may know them to be wrong are
asked to make allowances.

Chapter 8


Table 15. Ultimatum to the Bosnian Serbs.

First, what cards are in play and who holds them?

The United States controls the card, Bomb Serbs
into compliance. Such bombing would be done by
NATO, but this card is assigned to the United States
because, first, the United States has the necessary
military assets and, second, NATO as a whole, apart
from Britain, is prepared to support the United States
in this.


Confrontation Analysis

Russia has the important card, Back Serbs against

NATO. Russias pro-Slav sympathies and desire to
have a friendly Slavic nation controlling former
Yugoslavia incline it to play this card, if necessary, to
offset what it sees as Western anti-Serb bias.
The Bosnian Serbs have the same cards in this
confrontation as at the operational level (see table 15).
Finally, Britain has the card, Publicly oppose
bombing, a card it is inclined to play to keep a
reputation for impartiality and lessen the danger to its
vulnerable troops engaged in a humanitarian mission.
What positions are parties taking?
The United States and Britain (B) take somewhat
differing positions, but because their positions are
close, we have collapsed them into one. We have done
this by leaving it open whether, under the joint position,
B would publicly oppose bombing. B proposes to do
so (to show impartiality and protect its troops), and
the United States is willing to let it do so. The United
States, Britain, and the rest of NATO agree that the
Bosnian Serbs should cease fire and withdraw
weapons, obviating the need for the United States to
bomb the Serbs or for the Bosnian Serbs to retaliate
against UN personnel.
Where Britain differs from the United States and the
rest of NATO is in being opposed to the U.S. fallback
position, Bomb Serbs into compliance. Britains own
fallback position is to continue publicly to oppose
bombing. Britain takes this fallback position out of
preference (to show impartiality and protect its troops)
and to deter the United States from taking its fallback
position by making that position less comfortable for

Chapter 8


the United States (and so helping to give the United

States inducement and threat dilemmas that might
lead it to change its fallback position). Britain is not
taking this fallback position as a way of inducing the
Bosnian Serbs to accept the joint position U,B.
Russia (R) at this stage has not taken a clear position
on any issue except that it is against NATO bombing
of BS (it has agreed to close air support, but excluded
punitive air strikes). It has not taken the position of
backing the Serbs against NATO, but is implicitly
threatening to do so if NATO bombing is implemented.
Meanwhile, the United States, Britain, and the rest of
NATO take no position on whether Russia should
back the Serbs against NATO. That, they consider,
is up to Russia. But if Russia should carry out its
implicit threat to back the Serbs while NATO is
bombing them, there would be serious concern about
a deterioration of relations between Russia and the
United States, even fear of a new East-West
confrontation arising over the Balkans.
The threatened future, t, is that the United States
commences to bomb the Serbs into compliance (the
action would be taken by NATO, but our assumption
is that the decision would be taken by the United
States); Russia responds by backing the Serbs against
NATO, with all the geopolitical risks that entails; the
Bosnian Serbs respond by retaliating against UN
personnel; and the British publicly oppose bombing.
Whether or not it is credible, this is the future implied
by the parties current fallback positions.


Confrontation Analysis

General Positions
In table 16 we have made free use of the symbol ~ to
represent general positions (i.e., to show that a party
has not taken a clear position as to whether certain
cards should be played). The symbol ~ means that a
card may or may not be played. Hence, when
incorporated into a characters position, it means that
the character takes no position as to the playing of
that card.
We must use this symbol here in order to be realistic.
Your position is something you are willing to go to the
threatened future for. If you are not willing to go that far,
then at the moment of truth it becomes clear that
effectively you are not taking a position on that card
(i.e., your position is compatible both with it being played
and with it not being played). This is the case with the
U.S. attitude to Britain publicly opposing bombing; the
United States does not like it, but it is known it will accept
it. Similarly with the Bosnian Serb attitude to Russia
backing it against NATO; the Bosnian Serbs would like
it, but are not demanding it (at least not in this model,
although if we modeled Serb-Russian negotiations, we
might find the Bosnian Serbs demanding it). Russia has
not taken a firm position on any issue except NATO
bombing of the Serbs.
Use of the symbol ~ means that the different positions
of different parties may be compatible (e.g., in table
16 Russias position is compatible with both of the other
two, although they are incompatible with each other).

Chapter 8


Table 16. Grand strategic pressure on the Bosnian Serbs.


Confrontation Analysis

Linkages Between ModelsUse of Context Cards

We now have models of two confrontations, at the
operational and grand strategic levels, that are
clearly linked.
First, we can see that the reason the Bosnian Serbs
have the same cards in each model is that they are
playing simultaneously in both confrontations. At the
grand strategic level they are sending messages to
the United States, Russia, and Britain, while at the
operational level they are negotiating with the
UNPROFOR commanders staff and through it,
indirectly with the Bosnian government.
Second, it is clear that the U.S. fallback position in the
grand strategic game (Bomb the Serbs into
compliance) partly dictates the UNPROFOR
commanders fallback position at the operational level
(i.e., it means his fallback must contain the card, Call
air strikes against Serbs); however, it does not dictate
which other cards he must play. Because he is the
man on the spot, responsible for interacting with the
Bosnian Serbs and having immediate responsibility
for deciding if they have complied, he must decide at
what exact point this fallback position should be
implemented. Realistically, we should regard him as
holding and negotiating with the card, Call air strikes,
while knowing (and knowing that the Serbs know, etc.)
both that he has been directed to threaten its use and
that his decision to play it or not may be overridden by
a U.S. decision in the grand strategic game. Thus he
plays with this card knowing that at any time he may
be tapped on the shoulder and told what to do with it,
but fearing that if he cannot get agreement, he must
use it.

Chapter 8


The Bosnian Serbs tactic makes use of this. They

plan to negotiate with him while simultaneously
sending messages (both through him and through the
Russians) in the grand strategic game. Their aim is to
make the threatened future in that game so bad for
the United States that it will back off from its fallback
position, thereby making UNPROFOR back off from
its fallback position in the operational game. The
Bosnian Serbs plan to do this partly by threatening to
retaliate against UN personnel, partly by getting the
Russians to threaten to back them against NATO, and,
to a lesser extent, partly by the threat of British
opposition to bombing.
The UNPROFOR commanders tactic, on the other
hand, is driven by the fact that while opposed to
bombing, both personally and in line with his national
governments policy, to avert it he must credibly
threaten to do it. While credibly threatening, he can at
the same time use his discretionary powers to try to
water down the ultimatum (e.g., let the Serbs keep
their weapons in place provided he can monitor them),
thereby making conciliatory adjustments to his position
to make it more acceptable to the Serbs. He
desperately tries both tactics to get an agreed position
before the NATO ultimatum expires; however, his
bargaining position is strengthened by his ability to
use the nice-guy-mean-guy argument that, even
though he does not like bombing, he must order it if
the Serbs do not comply.
Thus the models are linked first, because the
Bosnian Serbs are playing in both games at once,
using the same cards, and secondly, the
UNPROFOR commanders choices in one game are
partly dictated by those made in the other, although


Confrontation Analysis

he does exercise some initiative as a subordinate;

that is, he has cards he can play or not play and can
choose his negotiating strategy.
What does the linkage do to characters preferences?
For example, to know if the United States and Britain
have a deterrence dilemma, we need to decide if the
Bosnian Serbs prefer t to U,B in table 16. Is this the
same as whether they prefer t to U, BG in table 15?
It is the same question, and will receive the same
answer, provided we follow our constant advice, which
is to consider relevant factors outside the model when
making assumptions within a model.
In answering the question relative to table 16, we need
to recall that in table 15 UNPROFOR is threatening to
play the card, Blame the Serbs, as part of t while at
the same time seeking to make compliance with U,
BG less onerous for the Serbs.
In answering the same question in the context of table
15, we need to recall that in table 16 the Russians are
indicating they will play the card, Back Serbs against
NATO, as part of t, and the British are indicating they
will publicly oppose bombing.
This consideration of external factors can be done
formally by adding context cards to a model. This
means adding cards below a line labeled Context,
as in the detailed model of table 7. We can add cards
showing the assumptions made about any external
factors. To show how this works, table 17 is the same
as table 15, but with added contextual assumptions
about what would be happening in table 16. These
assumptions are those stated above.

Chapter 8

Table 17. Ultimatum to the Bosnian Serbs with

context cards added from grand strategic model.



Confrontation Analysis

Dilemmas in the Grand Strategic and

Operational Dramas
We have looked at players positions and tactics in
the grand strategic and operational confrontations as
NATO presents its ultimatum to the Bosnian Serbs.
We have not spelled out the dilemmas they face.
The dilemmas in the grand strategic confrontation are
shown in table 18.
The United States and Britain face an inducement
dilemma and the United States faces a threat dilemma
in that t could lead to an EastWest confrontation, as
well as retaliation against UN personnel; therefore, it
seems that the United States and Britain would prefer
the Bosnian Serb position to t (this is their inducement
dilemma) and the United States would prefer not to
bomb the Serbs (its threat dilemma). Britains reactions
include fear of an EastWest confrontation (and of
endangering British troops on the ground); this fear
was conveyed to the Bosnian Serbs by David Owen,
who was moved by it to conduct diplomacy against air
strikes (Silber and Little, 1996, pp. 311-313; Owen,
1996, chap. 7) . The U.S. reaction included
rationalization of preference for t over column BS by
issuing an ultimatum and following the ultimatum,
arguing that NATO credibility is now at stake.
The United States and Britain face a deterrence
dilemma in relation to the Bosnian Serbs, who see an
EastWest confrontation in which Russia takes their
side (this is the BS view of t) as likely, in their view, to
end with the West making compromises in their favor,
and hence as being preferable to position U,B. U.S.
and British reactions include attempts to get Russia

Chapter 8


Table 18. Dilemmas in table 16.

to threaten the Bosnian Serbs with lack of support.

U.S. President Clinton phoned Russian President
Yeltsin and British Prime Minister Major visited Yeltsin
in Moscow to request Russia to pressure the Bosnian
Serbs to accept the position U,B.
The United States and Britain face a trust dilemma
because they cannot trust the Bosnian Serbs to
continue to implement the agreement U,B, even if they
should seem to comply (i.e., even if they begin to
implement it in the short term).
Russia faces a threat dilemma in that Russia, like the
United States and Britain, also fears a renewed East
West confrontation and would prefer not to play the
card, Back Serbs against NATO, as part of t. The
reaction included conflicting conciliatory and angry
signals from Moscow, ending, after a visit by British
Prime Minister to the Russian president, with
acceptance by Russia of a joint Anglo-Russian position
compatible with the U,B position in table 15 and a


Confrontation Analysis

threat to the Bosnian Serbs that if they rejected this,

Russia would not back them against NATO.
The next step in the resolution process is that Russia,
in an attempt to eliminate its threat dilemma, shifts its
position and brings about a new moment of truth;
however, before looking at this, we will look at the
dilemmas found in table 15, where the UNPROFOR
commander negotiates at the operational level with
the Bosnian Serbs. Table 17 is the same as table 15
except that context cards have been added to show
influences from the grand strategic level.
The dilemmas found at this level are shown in table 19.

Table 19. Dilemmas in tables 15 and 17.

U (UNPROFOR) and BG (the Bosnian government,

now sharing Us position) face deterrence and trust
dilemmas. Column t puts BS under no pressure to
accept column U,BG (this is the deterrence dilemma
facing U and BG.) In any case, because BS already
have formally accepted it, any further agreement on
their part to do so could hardly be trusted (the trust
dilemma facing U and BG). Reactions include U

Chapter 8


attempts to improve U,BG from the BS viewpoint by

reinterpreting the agreement to allow BS guns to stay
in their positions subject to monitoring. For BG, on
the other hand, these dilemmas are welcome as
arguments in favor of abandoning the position U,BG
and resorting to a position it much prefers (i.e., the
threatened future t ); therefore, BG urges strict
compliance with the agreement U,BG, hoping for noncompliance and the implementation of t.
U also has inducement and threat dilemmas. The
UNPROFOR commander himself openly states that
he prefers not to call air strikes against the Serbs, and
his dislike of this is supported by the grand strategic
fear of East-West conflict. The reaction includes a
mixture of anger and despair on the part of U, which
nevertheless keeps trying to get an agreement.
BG also faces a cooperation dilemma (it prefers not
to keep to the agreement) and a positioning dilemma
(it prefers BS to U,BG) because under BS it would
hope for eventual Western intervention. The reaction
is that these dilemmas are quite welcome for BG. It
has been forced into unwilling acceptance of column
U,BG. If this column is implemented, BG no doubt
intends to defect from it in due course, breaking the
cease-fire, perhaps gaining some ground, but
provoking further BS attacks on the basis of which it
can make further appeals to international opinion.

How Real Pressure Was Brought to Bear on

the Serbs
Note that in these models the Bosnian Serbs face no
dilemmas. Their position BS is preferred by them and
stable for others, so that they have no cooperation, trust,


Confrontation Analysis

or positioning dilemmas. Also, t is attractive to them

because it would involve Russian support, so that they
have no inducement or threat dilemmas. Also, they have
no deterrence dilemmas because t is feared by Russia
and the Western players, while BG likes BS.
All this changes after the British Prime Ministers visit
to Moscow and Russias subsequent agreement to
adopt a joint position with Britain and put pressure on
the Serbs to accept it. In table 20, the British-Russian
joint position is shown in column B,R. Here Russia
proposes to give its backing to the Serbs in return for
Serb agreement to cease fire and withdraw weapons,
while Britain publicly opposes bombing. Note that this
is compatible with the U.S. position. This is not to say
that the United States likes all its elements; it dislikes
Russian backing for the Serbs and British opposition
to bombing. Its compatibility with the U.S. position
merely means that the United States is prepared to
accept it, rather than go to its fallback position.
U, B, and R are now united against BS in demanding
that the Serbs cease fire and withdraw; however, the
significant point is that Russias fallback position has
shifted. If BS rejects B,R, then Russia no longer will
back the Serbs. This becomes clear to the Bosnian
Serbs when Russian Ambassador Churkin visits the
Serbs and suggests position B,R. On the same visit,
Russia assures the Bosnian Serbs of its willingness
to back the Serbs against NATO if the Serbs will agree
to position B,R by undertaking to send Russian troops
to supervise the withdrawal of Bosnian Serbs heavy
weapons. This changes everything for BS. It is now
implicit that, if they refuse this offer, Russia will not
back them against NATO. The threatened future t is
now one in which the world, including Russia, is against

Chapter 8


Table 20. Real pressure brought to

bear at last on the Bosnian Serbs.

them. Also, t now holds fewer fears for the United

States and Britain (no more fear of a new EastWest
confrontation). Britains fear of Bosnian Serb retaliation
against UN personnel remains, but affords little comfort
because U.S. troops are not threatened, and it is the
United States that controls the card, Bomb Serbs.


Confrontation Analysis

The dilemmas arising in table 20 are shown in table

21. The United States and Russia have only the
dilemma that they cannot trust the Bosnian Serbs to
keep a cease-fire for long, after they have agreed to
one. Britain has, in addition, the inducement dilemma
that it prefers the Bosnian Serbs position to t (because
of the threat t poses to British troops). This is a problem
for Britain, but does not help the Bosnian Serbs
because it is not Britains fallback position, but that of
the United States, which makes t unpleasant for the
Bosnian Serbs. The Bosnian Serbs have a deterrence
and an inducement dilemma of a kind that they can
see no way to solve.
The Bosnian Serbs response is to seize upon the offer
by B,R as a position it even prefers to its own because
it includes Russian backing. BS is joyful; arriving
Russian troops are greeted by cheering crowds.

Table 21. Dilemmas in table 20.

Chapter 8


There is now agreement by all parties on the position

B,R (the intersection of the two compatible positions
U and B,R). This is reflected in the operational drama,
as shown in table 22, where all parties now take the
same position. The remaining dilemmas are those of
cooperation and trust, indicated by the question marks
on the cease fire cards played by the Bosnian Serbs
and the Bosnian government. Note that the Bosnian
government here is accepting the position column that
is worst for it among the three columns shown. This
arose, as we have seen, because the Bosnian
government was threatened with a still worse column
(not shown here) where UNPROFOR would blame it
for the market square massacre; however, the fact
that it prefers either of the other two columns shown
to the one it is accepting emphasizes its cooperation
dilemma (and UNPROFORs trust dilemma). It cannot
be trusted to stick to the agreement. Reactions to these
dilemmas include the following: deceit on the part of
the Bosnian Serbs and the Bosnian government,
hiding their intention to break the cease-fire; disbelief
on the part of U, leading to the UNPROFOR
commander monitoring the cease-fire and trying to find
sanctions that will make defection unattractive and
benefits that will make the agreement attractive to the
Bosnian Serbs and the Bosnian government.


Confrontation Analysis

Table 22. Resolution (deceptive) of the

operational drama, with grand-strategic context.

Chapter 8


What This Analysis Has Achieved

In this chapter we analyzed the dilemmas faced by the
UNPROFOR commander and certain characters he
interacted with in 1994. We tried to show that their
reactions to these dilemmas can explain the emotions
felt and initiatives undertaken by the characters.
We explained how and why the UNPROFOR
commander, faced with Bosnian government
intransigence, reacted angrily, thereby making credible
a newly minted threat to place public blame on the
Muslims. This brought the Bosnian government into
fearful, apparent compliance with UNPROFORs
position. There remained the problem that the Serbs,
although they had declared themselves willing to
comply, now showed unequivocally by their actions,
public statements, and defiant emotional attitude that
their position now was one of non-compliance. They
were eventually brought into apparent compliance by
UNPROFORs threat of bombardment; however, this
threat was made credible only by grand strategic action
(the NATO ultimatum) and became adequate only
when Russia, again acting on the grand strategic level,
made it clear that the Serbs would not have its backing
against NATO. After it was made credible and
adequate, the threat achieved apparent Serb
compliance, although the Serbs remained defiant. The
Serbs took consolation from the fact that Russia,
thought to be their ally, had become involved. The
eventual attitudes of both Serbs and Muslims indicated
that their compliance was more apparent than real.
Our detailed explanations of these events are post
hoc; moreover, explanation by itself is of little value.


Confrontation Analysis

The important question is whether this analysis, if done

at the time, would have helped the UNPROFOR
commander achieve his mission objectives. This
question is addressed in the next chapter.

Summary of Chapter 8
To illustrate how the approach throws light on a realworld confrontation, this chapter reports an analysis
carried out for DERA of the episode in February 1994
when the UNPROFOR commander was tasked with
stopping the Bosnian-Serb bombardment of Sarajevo.
The analysis was drawn from published sources only.
Interviews were conducted with officers who had served
at various levels in the Bosnian mission, but not at this
stage or during this particular crisis. An explanation of
the behavior and reactions of the various parties is given
using drama-theoretic principles.

Chapter 9
Would This Analysis
Have Helped?

nalyzing a confrontation, as in the previous

chapters analysis of the Sarajevo crisis of
February 1994, does not answer the question, How
might confrontation analysis have helped the
commanders operation-level decisions, if the method
had been available to him at the time?
To answer this, we will need to assume that our
analysis is broadly correct, despite the fact that it is
based on published accounts written after the event,
rather than on participants on-the-spot judgments. Not
only are published accounts a poor guide to how things
seemed at the time, they also have the disadvantage
of concentrating on events at the operational and grand
strategic level, rather than the tactical level. The effect
is that we will not be able to say much about how the
commander might have devolved his operational
strategy to lower levels.
Despite these difficulties, we will project the analysis
backward, beginning with the situation after the market
square bomb, and imagine what might have happened
if the commander had been able to formulate and
follow a confrontation strategy guided by our analysis.



Confrontation Analysis

Rerunning the History of a Crisis

On February 5, 1994, let the commander begin by
building the model in table 13. At this point, he has
realized that the Bosnian president will not allow his
officials to attend cease-fire talks because he is trying
to use the situation to bring about international
intervention on his side.

Determining End States

In the course of building this model, the commander
has defined his end state. This is column U, the future
in which BS (Bosnian Serbs) ceases fire and withdraws
weapons, BG (Bosnian government) also ceases fire,
U (UNPROFOR) remains impartial (not blaming Serbs
nor calling air strikes), and there is no retaliation
against UN personnel.
This is the same as Us position, although a characters
end state is not always its position. Your position is your
overtly declared end state, what you propose to the other
characters as a solution. We have seen that a character
with a cooperation dilemma has a temptation (perhaps
working with others) to move from its position, if
accepted, to another future that it prefers. It may solve
this dilemma by deceit, so that its secret intention and
actual end state is to carry out its temptation, not its
position. Again, a character may have reluctantly
accepted a position while planning for it to be rejected
in favor of another. In this case, the latter rather than
the former will be its actual end state.
In table 13, it appears that each partys end state is
identifiable with its position; thus, the commander, in
building this model, has determined not only his own

Chapter 9


end state, but the other parties too. He has identified

their goals.
End states found in a model of a common reference
frame, which is what a card-table is, are necessarily
simple. Simplicity is desirable. It is also realistic. In a
confrontation, parties end states, representing what
they want others to accept, are necessarily simple.
At the same time, the commanders staff will look at
these end states in more detail to explore their
implications and consequences. They do this by
adding extra cards, both internal (i.e., above the
context line) and external (below it), as in table 7.
Simplicity, clarity, and certainty are lost in this process,
but important points may be uncovered and difficult
questions raised. Going into detail in this way is also
necessary to devolve a strategy to the tactical level.
To find end states for a mission in a card-table model,
you obviously must model that mission, and the
mission must be a confrontation (i.e., an Operation
Other than War [OOTW]). In our example, the
commander has modeled the mission he has just
received, not his previous and continuing humanitarian
mission, and has determined his end state for this new,
obviously confrontational mission. Lower-level,
devolved models will enable determination of end
states for lower-level missions.
Conversely, to determine his superiors end state, and
the end state of his superiors superior, the commander
models the grand strategic drama by building the
model in table 16. In doing so, he identifies his
superiors intent by modeling how Britain (his country),
the United States (supported in this by the rest of
NATO), and Russia see the application of pressure to


Confrontation Analysis

the Bosnian Serbs at the grand strategic level. Observe

that his superiors intent appears, not as a clearly
defined mission, but as a position in a political drama
being played out between national governments. The
end state of this conglomerate superior is column U,B
in table 16.
The commanders end state is, in general, a more
detailed implementation of his superiors end state,
just as the end states in strategies devolved from his
strategy will be more detailed implementations of his
own end state. In moving directly to achieve his own
end state he is helping his superiors to achieve theirs
and using his subordinates to implement matters of
detail within his.
In the case we are considering, the commander finds
a problem with his superiors model of the situation.
They are implicitly assuming that the Bosnian
government shares the position of the U.S. and British
governments, or at least accepts what these
governments think is good for it. BG (the Bosnian
government) is not even a player in the grand strategic
model. This reflects the way the governments see this
moment of truth. They see themselves as negotiating
with each other over how to bring the Bosnian Serbs
into line. BG does not enter into it.
How should the commander deal with the fact that,
contrary to his superiors assumptions, BG is planning
to take advantage of their blindness by refusing to
settle, knowing that the Serbs will be blamed for the
resultant perpetuation of the bombardment?
In light of his superiors intent, the answer is clear. He
must use his own initiative to bring BG into line.

Chapter 9


Centers of Gravity and Sequences of

Decisive Points
Use of some familiar military terminology may help
describe what the commander does. We can say he
attains his end state by finding a sequence of decisive
points by which to destroy the centers of gravity of
non-compliant parties.
The significance of looking for other parties centers
of gravity is this. Chapter 7 contained a description of
the commander as cycling through a number of
dilemma eliminations to achieve his end state;
however, formal analysis of dilemmas is never enough.
It is essential to look at the real-world context in which
dilemmas arise to see how to eliminate them in the
desired manner (i.e., without pressuring others into
further escalation or into adoption of yet more contrary
positions). This can be described as a matter of
defining and attacking the center of gravity of their
As an example, the center of gravity of the Bosnian
governments non-compliance could be defined as its
project of working on Western public opinion to bring
the West to its side. This center of gravity was effectively
attacked by the UNPROFOR commanders threat to
blame the Muslims for the market-square bomb.
This illustrates the point that to define a non-compliant
partys center of gravity, it is helpful to first decide what
its policy is. Is its position non-compliant? If so, it must
have a policy by which it hopes to get acceptance of
that position, and that policy must be overthrown, as
the UNPROFOR commander overthrew the Bosnian
government policy of prolonging the Serb bombardment


Confrontation Analysis

to arouse international opinion against the Serbs. On

the other hand, it may be that its overt position is
compliant, and its non-compliance lies in its presumed
intention to defect from it. In that case, it is the policy of
defection that must be overthrown. Its center of gravity
may be the belief that defection will be possible or the
belief that it will be profitable.
In seeking to define centers of gravity, it usually is
necessary to analyze (formally or informally) the
internal drama going on within other players. The
present policy of non-compliance has been arrived at
as the resolution of a confrontation between various
internal subcharacters. Our aim must be to launch an
attack on certain agreed underpinnings of that policy,
thereby throwing it into crisis and bringing about a shift
to a policy of compliance. The commanders task is to
eliminate dilemmas at his own level, working where
necessary through subordinates, coordinating with
others on his own level, and requesting assistance as
necessary from higher levels.
Each round of dilemma elimination at his own level
may be described as a decisive point for the
commander. His confrontation strategy as a whole may
be described as a plan to move through a sequence
of decisive points, success at all of which will eliminate
non-compliant centers of gravity and bring all parties
into sincere compliance, so achieving his end state.

Planning and Following a Confrontation Strategy

To show how he proceeds, we will describe several
cycles of dilemma elimination. These will illustrate both
how he might have planned a confrontation strategy
and how he might have carried it out.

Chapter 9


Step 1: We have said that he starts by building table

13. The dilemmas found here are set out in table 14.
BS has no dilemmas. Both U and BG have deterrence
dilemmas and trust dilemmas. Neither BS nor BG is
under pressure to accept the commanders position.
Step 2: The commander must solve his deterrence
dilemma against at least one of the other parties (i.e.,
bring pressure on at least one). Now he knows that
his superiors are thinking how to bring BS (the Bosnian
Serbs) into line. They are not even thinking about BG;
therefore, he decides that he must deal with BG.
Now BGs deterrence dilemma means that its position
is weak at the operational level because it has no
means of inducing the commander or BS to accept its
positive position (column BG). Its strategy is to remedy
this by operating on the grand strategic level to change
the commanders preferences from above. He can
forestall this if he can overcome his deterrence
dilemma toward BG using a credible, useful threat that
will induce it to comply. (He may not be able to bring it
into true compliance because it may still intend to
defect from his position, but apparent compliance is a
first step.)
The commander brainstorms with his staff for
possible cards. The commander considers the BG
center of gravity is its desire to swing international
opinion on its side; therefore, an appropriate card to
pressure BG with is, Blame Muslims for market
square bomb, a card suggested by the evidence
that they were responsible. He provisionally adds
this card to his hand.


Confrontation Analysis

Having done so, he builds table 11, which includes

this new card in his fallback position. He finds, on
analyzing this table, that he has solved his deterrence
dilemma but given himself inducement and threat
dilemmas (in that he prefers not to use the new card).
He decides to overcome them either by showing anger
(as was actually done in 1994) or in some other way
found by brainstorming and role-playing the problem
(e.g., he might represent Muslim blameworthiness as
a logical deduction from BGs unwillingness to adopt
the position U).
Solving his own dilemmas will, he decides, heighten
BGs inducement dilemma to the point where it can
solve it only by accepting his position.
Step 3: At this step he successfully carries out, or, in
planning mode, rehearses, these dilemmaeliminations. Rehearsal methods described in chapter
10 are a good way of testing this phase of a
confrontation strategy.
Step 4 consists of a return to Step 1. Because the
planned dilemma-elimination has succeeded (or, in
planning mode, has been assumed to have
succeeded), in shifting BGs position to overt, if
unreliable, acceptance of his position, the commander
builds the model in table 17.
This model has context added from the grand strategic
drama because this now seems to be of overriding
importance. To establish this context, the commander
also builds table 16, thereby modeling the problem
faced by his superior.
In the new model at his own operational level, he faces
up to the problem of bringing BS into apparent

Chapter 9


compliance. He identifies the dilemmas in the model

as being those in table 19. He finds that he faces
dilemmas of deterrence, inducement, and threat, as
well as trust. BS has no dilemmas. BG has dilemmas,
but because BG is still clearly aiming for a position
other than the one it is overtly taking, it welcomes them,
regarding them as arguments against its position and
in favor of the position it would rather be taking.
Step 2: What can the commander do? In neither the
operational nor the grand-strategic level table does
BS now face any dilemmas. No pressure is being
brought on BS.
This is disconcerting. The commander decides to use
brainstorming and role-playing to search for the BS
center of gravity. He finds that BS does not fear t in
either model because BS sees R (Russia) backing it if
NATO attacks. It seems, then, that BSs belief in
Russian backing is the BS center of gravity. From his
superiors model in table 16, the commander finds that
R has an inducement dilemma and a threat dilemma
in backing BS in the event of NATO air attacks.
That is useful information. The commander decides
to use it to attack the BS center of gravity (after
checking with his superiors, which he needs to do
because the tactic of using these Russian dilemmas
has grand strategic implications). His plan is to point
out to BS that R will certainly not back it when the
chips are down. Getting this message across
effectively should eliminate his (UNPROFORs)
deterrence dilemma and give BS an inducement
dilemma by making BS prefer his position to t.


Confrontation Analysis

Simultaneously, he plans to eliminate his own

inducement and threat dilemmas and give BS a
deterrence dilemma by pointing out that NATO is
committed by its own ultimatum, and cannot now back
down. NATO, unlike the UN, cannot afford to lose
credibility. NATOs commitment is such that even the
threat of retaliation against UN personnel is not
sufficient to make column BS preferred to t; such a
threat merely angers Western public opinion and
increases the U.S. preference for t over the Bosnian
Serbs position, and, he will point out, U.S. preference,
not British, is what makes t credible. He plans to point
out that Russia knows all this, and therefore will itself
back down; this is why it has not even taken a firm
position. He will point out the inevitable conclusion
that Bosnian Serbs intransigence will lead to NATO
air action plus withdrawal of Russian support. He will
try to get this message reinforced by all sources in
contact with the Bosnian Serbs, thus achieving unity
of effort. For good measure, to take away any crumb
of comfort the Bosnian Serbs may find in the
threatened future t, he will add the assurance that while
Britain does at present publicly oppose bombing, it
will not do so if NATO takes action.
The commander builds these plans from analysis of
dilemmas in conjunction with appreciation of the
attitudes of the Bosnian Serbs. He has one problem,
though. To overcome his inducement dilemma, he
must find some way of dealing with his own reluctance
to use air strikes.
He does not like air strikes. He fears their damaging
effect on his ongoing humanitarian mission and prefers
not to risk the lives of UN personnel. Yet his analysis

Chapter 9


makes it clear that to get rid of his inducement dilemma

and give the Bosnian Serbs a decisive deterrence
dilemma he must find some way of preferring t to the
Bosnian Serbs position. He finds an answer to this
dilemma in the principle that he must try to implement
his superiors intent. His superior, in this case, is a
congeries of governments swayed by media-hyped
public opinion and arriving at decisions through a
process of political confrontation. Right or wrong, that
is his superior. It is not, in this case, manifestly wrong.
It is clear that the firm line being taken by the United
States and supported by the rest of NATO (apart from
his own government) means that his superior is
swinging in the direction of maximum pressure to make
the Bosnian Serbs comply; therefore, clearly his
effective preference, mandated by his superiors intent,
is for t rather than BS, much as he dislikes t.
BS simply must give in.
Step 3: BS does give in (or its role-players do, if we
are still in planning mode). Another decisive point is
passed. This occurs because U convinces BS that
the true model it faces if it does not shift its position is
that in table 23. Here, t involves neither R backing the
Serbs nor B publicly opposing bombing, but continuing
air strikes preferred by U to BS.


Confrontation Analysis

Table 23. Ultimatum to the Bosnian Serbs

with threatening grand strategic context.

Chapter 9


The Rerun Compared to the Reality

Observe that this rerun of the past achieves the
same end result as the reality, but achieves it in a
different way.
In reality, U got BG to overtly accept Us position by
threatening to blame the Muslims for the market square
bomb. This achieved the first round of dilemma
elimination described in our rerun; however, rather
than being done in a planned and thoughtful way, with
an adequate appreciation of risks and ramifications,
the elimination was done impulsively and emotionally.
The commander got angry and made his threat
credible in that way.
The commanders instincts were right, and he was
successful; moreover, when there is little time to act,
action must be fast and decisive. As a whole, U
(including the commander, his staff, and UNPROFOR
coevals) would have benefited from a clearer, more
communicable understanding of what he was doing
to win a confrontational battle. A major benefit might
have been to win in such a way as to enlist more willing
support from BG for Us position (e.g., by making it
seem to BG that Us motivation was to prevent BG
from ruining its case at the court of Western public
opinion). BGs actual support, after being coerced by
angry threats, was unwilling and reluctant.
The second round of dilemma elimination was not
achieved at all; not, that is, at Us operational level. U
negotiated with BS throughout the 10 days of the
NATO ultimatum in terms of the model in table 17 and
did not succeed in eliminating any dilemmas from it.


Confrontation Analysis

This was the model U had created by successfully

bringing BG into line. In it, BS faced no dilemmas,
and hence U could bring no pressure to bear on BS.
Aware of this, U tried to fend off air-strikes by
concessions to BS in terms of the locations to which
heavy weapons must be withdrawn and the degree of
their supervision, hoping thereby to entice BS into
withdrawal. In this way U tried to solve its deterrence
dilemma and give BS an inducement dilemma by
sweetening Us offer. Apparently U failed in this, as
BS simply failed to withdraw its weapons (Silber and
Little, 1996, pp. 316-317).
U was saved, according to our analysis, by
simultaneous action on the grand strategic level.
Effectively, the politicians came to Us rescue. After
British Prime Minister Majors visit to Moscow, Russian
Ambassador Churkin saw the Bosnian Serbs at their
headquarters in Pale, told them that Russia backed
the NATO ultimatum, and offered them Russian
backing if they accepted its conditions. He made this
last offer by saying he would send Russian troops to
monitor the withdrawal of BS weapons. Note that he
did not explicitly tell BS that if they rejected NATOs
conditions, then Russia would leave them in the lurch.
It is undiplomatic to spell out your threats. To get willing
cooperation, a positive note is necessary. In failing to
tell them that Russia would, of course, back them
under all circumstances, the ambassador was being
sufficiently explicit and at the same time positive, as
the Serbs made clear by joyously accepting his
conditional offer of support.
Such was the reality. In our rerun, U takes advantage
of precisely the same configuration of grand strategic
dilemmas to solve the problem at operational level,

Chapter 9


although note that in doing so, it first checks its strategy

with its grand strategic superiors. Checking its strategy
in this way also enables it to achieve useful
coordination with the grand strategic level, giving
reassurance that the messages BS receives on that
level will confirm what U is telling it, the essential point
of which is that it cannot rely on Russian backing
against NATO air strikes. Note that Us argument that
BS cannot rely on R depends on the very Russian
inducement and threat dilemma (that R feared a t that
might lead to a new confrontation between East and
West) that led Ambassador Churkin to visit Pale and
make the same point.
In our rerun, U uses these powerful arguments to win
the confrontation at operational level. BS gives in. We
cannot, of course, say with certainty that U could have
achieved this. BS might have persisted in its grand
strategic efforts until R communicated directly with it.
What we can say is that if U attempted dilemma
elimination along the lines sketched in our rerun, it
would have been pursuing a confrontation strategy
coordinated between grand strategic and operational
levels, and thereby would have been pursuing its
superiors intent. It is certain, too, that such a strategy,
coordinated between different levels, would have been
far more effective than the ineffective and ultimately
damaging efforts to sweeten its offer to BS that were
pursued in reality.
In military terms, Us strategy would have seized and
maintained the initiative by attacking BSs center of
gravity rather than, as in reality, trying unsuccessfully
to conciliate BS by progressively abandoning Us
own position.


Confrontation Analysis

The Next Step in Our Rerun of Reality

Stabilizing the Agreement
Step 4 (A New Step 1): Assuming we have coerced
BS to overtly comply with its position, U returns to the
drawing board. Its new model of the situation is that
seen in table 22. The dilemmas it finds in this are set
out in table 24. BS and BG both have temptations to
defect, preferring to break the cease-fire not only
individually but collectively. They would prefer, in other
words, to resume hostilities. BS would prefer this
because they wish to gain ground while they can, being
aware of the growing strength of a CroatMuslim
coalition being formed against them. BG would prefer
it, now or in the near future, partly because they hope
to be able to drive the Serbs back and partly because,
if they fail, the renewed bombardment of Sarajevo
might swing world opinion in favor of intervention on
their behalf.
The agreement is holding for the time being because
it has just been signed and UN troops have been
interposed between the two sides. In these
circumstances, neither side wants to be seen to be
the first to break it; nevertheless, it requires stabilizing
before one or the other moves against it.
Step 2: How is the agreement to be stabilized? The
problem is to change the perception by BS and BG of
the advantage to them of an end to the cease-fire.
Unfortunately published sources give too little
information for us to say how this problem can be (or
was) tackled. Threatened media exposure of violations
may deter BG but not BS, whose image in the media
can hardly get worse. Credible threats of reprisals may
be needed to deter BS. Recognizing that the dilemmas

Chapter 9


Table 24. Dilemmas facing characters in table 22.

facing BS and BG are cooperation dilemmas, for which

a positive tone is appropriate, U makes sure that all
threats are part of a harmonious, positive stabilization
package constructed with full consultation and
including carrots, however symbolic, as well as sticks.
This package must be implemented at the tactical
as well as operational level. Consequently the
commanders staff, after putting all the detail it can
think of into an expansion of the overall model, draws
up guidelines for local commanders. Within these
guidelines, local commanders can analyze their
confrontations and draw up local packages to
eliminate dilemmas in them. As in the local
commanders model in table 3, BS and U units often
will take differing positions in these local models
(e.g., in regard to whether certain arrangements
conform to the agreement). Negative emotions then
will be demanded, within the overall positive tone of
the agreement.
There are potential problems of coalition relations in
regard to Russian troops guarding some BS weapon
cantonments. These will have to be managed in a


Confrontation Analysis

simultaneous, coordinated manner at operational and

tactical level.
Step 3: Suppose their dilemmas of cooperation and
ours of trust are solved, either in reality or role-playing
mode. The agreement is then stable, for now;
however, it is likely that external events, interruptions,
may disturb it by changing the conditions under which
it holds.

Predictions Made by Drama Theory and

Confrontation Analysis
Rerunning reality as we have done highlights the kind
of predictions drama theory makes. It is important to
understand that they are not deterministic. Neither are
they probabilistic.
A character may react in various ways to the dilemmas
it faces; however, if the assumptions fed in are correct,
the analysis identifies the dilemmas each character
must react to and allows us to predict that it will react
in such a way as to try and eliminate those dilemmas.
What reactions are available to it will depend on the
concrete details of the particular situation (i.e., on the
friction it encounters in trying to change its moment of
truth in one way or another). Ways of changing the
moment of truth may include changing position,
changing preferences, thinking up new cards to play
or taking action (e.g., military strikes) to deprive others
of cards.
In this way, the theory uses models in an unusual way.
We cannot, as in most other approaches, build a
model, manipulate it formally to derive predictions, then
come back to reality to apply those predictions. From

Chapter 9


our model-building and formal analysis, we derive not

predictions, but questions about reality in these forms:
What means are available for X to try and solve suchand-such dilemmas? It is our answers to these
questions that provide us with the predictions we feel
able to make.
Use of a confrontation model, rather than turning us
away from reality, turns us back to it with searching
questions to be answered.

Comments on This Rerun

This particular rerun of reality has suffered from being
based on assumptions taken from third-party,
retrospective, published sources, rather than the views
of actual participants at the time. This affects more
than the reliability of the analysis, although we have
asked readers to discount inaccuracies. Drama theory
is based on the game as perceived by the players.
Their views, not those of others, determine their actions
and reactions; therefore, theoretically it needs to be
based on their views at the time.
Relying on published sources also means that we have
incomplete data, particularly as regards confrontations
at the tactical level. We could not say much about
how BS and BG might have been brought into actual,
not just apparent, compliance with a cease-fire.
Accepting all this, what difference might it have made
if a confrontation strategy had been formed and used?
It appears that the same result would have been
achieved in bringing the BS and BG into apparent
compliance with a cease-fire, with certain differences.


Confrontation Analysis

The bringing of BG into apparent compliance might

have been done in a planned way rather than through
spontaneous reactions. Further, it might have been
planned to make BG more willing to comply, rather
than feeling coerced.
The bringing of BS into apparent compliance might have
been achieved sooner on the operational level instead
of being achieved, after unavailing concessions had
been made, on the grand strategic level. In achieving
this result on the operational level, the commander
would have been implementing his superiors intent.
Some of the tactics used, in particular, the forecasting
of Russian reactions, might have had to be checked
with his superiors. Others would not. Coordination
between the operational and grand strategic levels
would have been enhanced.
Achieving the result on the operational level, in the
manner outlined, would have meant some differences
in the result on the grand strategic level. The actual
result involved Russia. If the commander had achieved
it in the manner we discuss, Russian involvement
might not have been necessary. Perhaps it was
desirable. Perhaps Russian involvement was a grand
strategic objective. This emphasizes the necessity for
a commander to check with his superior before using
tactics with higher-level implications. His checking
might have led to an improved plan for obtaining
Russian involvement, if this was an objective.
However much or little the putative result would have
differed from the actual one, it would have been
achieved (if a confrontation strategy had been used,
at the initiative of the commander) instead of by lastminute action at a higher level after the commander

Chapter 9


had made significant concessions from his original end

state. It would have enabled the commander, from
the beginning, to seize and maintain the initiative.
Insufficient data in public sources make it hard to say
much about turning apparent compliance by the
Bosnian parties into actual compliance. It may or may
not have been achievable using available assets and
within given guidelines. If it was not, attempts to
formulate a confrontation strategy would have revealed
this, enabling inadequacies to be reported upward and
requests for extra support clearly justified.

Summary of Chapter 9
The analysis reported in the previous chapter is used
to show how in a real-world case the commander in
an OOTW might have formulated and implemented a
confrontation strategy. Two rounds of dilemma
eliminations that might have been carried out as part
of such a strategy are suggested.
Comparing this rerun with what actually happened,
the conclusion is that the first round of dilemma
eliminations might have been achieved much as in
reality, but in a more planned way and perhaps with
greater effectiveness in inducing willing compliance.
The second suggested round of dilemma eliminations
was not achieved at all by the commander on the
operational level; it required grand strategic action. If
it had been carried out by the commander, he would
have been operating a strategy coordinated between
operational and grand strategic levels, seizing and
maintaining the initiative, avoiding costly concessions,
and carrying out his superiors intent.


Confrontation Analysis

While we hope this attempted rerun may be useful, it

should be pointed out that it makes use of hindsight,
whereas confrontation analysis is most effective when
used to understand and plan for an actual, ongoing
operation, not one that already has taken place. This
is so because various futures that seemed only too
possible at the time seem by the exercise of hindsight
to have small probability just because they did not
happen. This diminishes the effectiveness and impact
of confrontation analysis. Use of hindsight also makes
the rerun questionable, if not useless, as a test of the
theory or technique.

Chapter 10
Immersive Briefing and
Mediation Support

he main topic of this book has been how to build

and use a confrontation strategy to win an
Operation Other Than War (OOTW).
In dealing with this, we have several times mentioned
two related topics we now discuss in more detail.
The first is role-playing. We will discuss a method of
writing briefings for role-players called immersive
briefing. This is a drama-theoretic method of looking
at a set of linked confrontations from the viewpoint of
each party, thinking how they must see the situation
and how they must see the way each other must see
it, and so forth, and basing on this a set of briefings
for role-players to take the parts of real-life characters.
The second topic is mediation support. Generally we
have assumed that confrontation analysis is used for
unilateral decision support, helping one party deal with
others. A commander would use confrontation analysis
in this mode to bring external players into compliance
with his end state. But, a commander also must deal
with confrontations that are essentially cooperative,
in that the other players are on his side and his
superiors intent is that they should all coordinate their
activities to achieve given objectives. Mediation
between the parties is then more appropriate than



Confrontation Analysis

support to just one player. For this we will discuss a

modified form of confrontation analysis.

Immersive Briefing
In step 2 of a confrontation strategy, you decide on a
batch of dilemmas to eliminate and think how to do it.
In step 3 you carry out your elimination plan, or if you
are planning rather than implementing a strategy, you
think through what the results are likely to be. From
there you go on to the next cycle, modeling the new
situation you have brought about or imagined.
This was the process we described in chapter 7 and gave
an example of in chapter 9. However, when planning or
checking your strategy, you can do better than just
thinking through what its results are likely to be.
There is an inherent difficulty involved. On going into
a moment of truth you open up your beliefs and values
to being changed; hence, it is theoretically impossible
to be sure how you will come out of it. While you cannot
be sure, an excellent way of stretching your
imagination and letting you feel the forces of change
beforehand is role-playing.
Role-playing is widely used for this purpose. Its results
generally are seen by good actors as abysmal. The
reason is simple: however good you are, you cannot
do much with a bad script. Having a good script is not,
however, a matter of having words to learn. It is a
matter of knowing the life situation of the character
you are acting, meaning not its personality or
character, but what it is trying to achieve, and why
and how, and what it thinks others are trying to achieve,
and why and how. Knowing these things about your

Chapter 10


character, you can start to act, not by putting on

another persons mannerisms, but by putting on its
life, playing the game it has to play.
This requirement of good acting is precisely the
requirement we have for exploring what may happen
at a moment of truth. We need to be able to experience
the moment of truth as closely as we can, and so we
need to recreate the characters we will be playing
against. If we can be helped to throw ourselves
completely into the problems and viewpoints that other
characters inhabit, we can recreate them as we want
them, try out things with them, and have the added
bonus of a deeper insight into their viewpoint. In this
way we can check assumptions made in the course
of planning a strategy (e.g., whether a non-compliant
character, placed in a certain situation, has any
alternative other than to give in).
Immersive briefing is a drama-theoretic way of
producing briefings for role-players. Unlike most roleplaying methods, it tends to produce good acting. Its
effectiveness derives from its analytical basis. It puts
participants into the positions of other characters, so
getting them to understand their subjective feelings
and perspectives and enabling them to come up with
creative ideas (e.g., this character might react in suchand-such a way). Participants also come to understand
their own confrontation strategy on a deep, intuitive
level, enabling them to criticize it and suggest


Confrontation Analysis

What Comprises a Briefing

An immersive briefing is constructed using special
software. With its aid you create a plot situation where
a number of role-players can be briefed to interact,
without scripts, as characters in a set of interconnected
confrontations. There are no rules and no points to
earn. The method is based on confrontation analysis
of a situation, but briefings can be given to and used
by role-players with no knowledge of drama theory or
confrontation analysis.
Separate briefings are given to each role-player, who
may be an individual or a team. A character in the
confrontation that is being role-played is generally an
organization (e.g., a country or a an ethnic group).
Individual role-players are told to act as responsible
representatives of their organization. The briefings they
are given resemble those given to a commander taking
over tasks from a predecessor. They should be both
complete and concise. A role-player taking on a
character is briefed on the following items:
Character background, including its organization,
the internal subcharacters and subconfrontations
in its organization, and its relevant history
Values and motivation, recognizing that a
character will have many kinds of value systems,
selfish and unselfish, long- and short-term, and
that these often conflict
Current projects (current goals)
Current relationships with other characters
Other characters backgrounds, values, and
projects and the relationships between them

Chapter 10


Confrontations it is involved in with other

The logic of this briefing is that a characters
background and relationships explain its values, its
values explain its projects, and its projects, because
they are interdependent with the projects of others,
explain the confrontations it is involved in.
In relation to each confrontation, a character is told
the following facts:
Who is involved (characters)
What each character can do about it (cards)
What position each is taking (i.e., what solution
to the conflict each recommends)
What fallback position each character says it has
in case the others will not accept its position (or
it believes will not genuinely accept it)
What future threatens if these fallback positions
are implemented
What future is forecast if currently implemented
policies continue. This may or may not be the
same as the threatened future
What strategy it and the other characters
currently are pursuing as each tries to get the
others to accept its position.
A character may be role-played by a team of up to
about five people who digest and discuss their
characters briefing among themselves. During internal
discussions, these teams are segregated in separate
rooms, in separate corners of a large room, or even in


Confrontation Analysis

separate locations, as when the drama is played over

a computer network.
After receiving and digesting their briefings, characters
(as played in role by role-players) are allowed to
communicate with each other to implement their
strategies (i.e., so that each character can try to get
others sincerely to accept its position for
implementation). For this each character tries to make
credible the threats (often implicit) and promises
specified in its strategy. Communication may be through
arranged meetings, fixed by intermediaries or by e-mail
over a network. Various communication media may be
used. Communication by public announcement (e.g.,
giving a press conference), knowing that another
character will read about it, can be simulated, such as
by sending out a general e-mail purporting to be a news
report of a press conference. Internal discussions
continue between meetings.
Records of communications between characters help
in later analysis. They can be kept automatically if
communication is over a network.

What an Immersive Briefing Represents

An immersive briefing really represents a particular
characters memory, or at least, the part of it that is
relevant to a particular set of confrontations. If a
character is an organization, its briefing represents its
organizational memory. That is why a briefing may be
likened to the briefing given a new commander when
he takes over responsibilities from a predecessor.
Three things follow from the fact that a briefing
represents a characters memory:

Chapter 10


A character can consult its briefing at any time.

As a situation develops into a new one, a
briefing becomes out-of-date. Like a diary that
stops at a certain date, it records things as
they were, or seemed to be, at a point now in
the past. Replacement of the briefing by a
more up-to-date one is possible. Alternatively,
characters can be left to make their own notes
of new developments.
Briefings are subjective and differ from one
another. Character A may be briefed differently
from B about past events, about its own or
anothers position in a particular conflict, and so
forth. Certainly characters briefings will differ in
the values reflected in them, because each
briefing will reflect the values of the character
being briefed. Consequently briefings are
second- and third-order subjective, as well as
first-order; that is, characters are briefed as to
(their beliefs) about others beliefs and values,
others beliefs about their own and others
beliefs, and so forth (see Bennett, 1977).

Using Software
The package for authoring an immersive briefing
produces a suite of software briefings, one for each
character, reflecting the characters different
viewpoints and information. Current software provides
briefing information both in text form and card-table
form. Text or card-table is called up in a computer
screen window by clicking on menu items or pictures
of the characters and the confrontations between
them. Further clicking on different parts of a card-table


Confrontation Analysis

calls up textual information about characters, positions,

fallbacks, and strategies in that confrontation. Figure
3 shows how a briefing written for the character,
Bosnian Govt, looks before any pictures or menu
items have been clicked. This briefing was one of an
experimental suite of briefings written for DERA.
Pointing and clicking through an immersive software
briefing is, even without role-playing, an effective way
of absorbing a confrontation analysis. It might also be
used, therefore, to present and distribute analyses.
Users could brief themselves by looking through the
briefings given to different characters and comparing
their viewpoints.

Friction and Change in Immersive Role-Play

Immersive briefings generally are designed to bring
characters to a moment of truth. Recall what this means.
A moment of truth is a point at the climax of a
confrontation when characters achieve the following:
They finally understand each other (or think they
do), including each others positions and
fallback positions.
They are, as a result, brought up against the
need to change their fixed views of themselves,
their situation, and each other if they are to avoid
falling into conflict.
At such a point characters tend to change their
definition of the confrontation they are in by redefining
the characters in it, the cards they can play, and their
preferences between futures, as well as their positions;
however, their ability to do so is limited by the weight
of existing values, evidence, mission definition, and

Chapter 10

Figure 3. Computer screen showing briefing

on character role for Bosnian government.


Confrontation Analysis

so forth (i.e., by friction). Emotion enables them to

break out of existing assumptions and routine
reactions, and so to overcome friction.
Role-players accordingly are told that the frames and
positions their briefings give them are initial frames
and positions only. They have inherited them from their
predecessors but they can change them. In this
respect, immersive role-playing differs from gameplaying or ordinary role-playing. The only constraints
are imposed by the following circumstances:
The past, consisting of characters backgrounds
and the events and projects that brought them to
their current impasse
Characters expectations of each other, and the
difficulty they may find in changing these. It will
be of little use to me to undergo a conversion so
that I prefer to carry out a promise or threat if
you continue to believe I will not. Equally, a
change of position or discovery of a new option
not listed in the card-table will need to be
communicated convincingly
Role-players sense of responsibility for properly
representing their character. An immersive roleplay can, in theory, be spoiled by an
irresponsible player who acts out of role. At the
other extreme, role-players may need to be
reminded that they are free to change, as some
are over-responsible. Such reminders can be
given in debriefing sessions arranged to follow
each round of negotiations.
Over-responsibility for the reality of your character is
a particular problem in role-playing past real-life

Chapter 10


events, as players tend to think they should do as their

real-life counterparts did; therefore, it is best to
fictionalize past events, at least to the extent of
changing characters names.
The degree of friction a character meets should
depend on its briefing, which tells it how important
certain values are to it and which facts it sees as
incontrovertible, and on the pressures brought to bear
on it in interactions with others. Emotions engendered
in a role-player may enable it to overcome even
considerable friction to reframe the situation. Roleplayers should experience the fundamental paradox
of drama theory (i.e., the more I feel hemmed in by
fixed constraints, the more emotion I feel, hence the
more I am inspired to think of changes, so that the
more unchangeable a situation feels, the more
changeable it becomes). The realism of this
experience is enhanced if players first run a role-play
under instructions that they should stick strictly to their
briefings. Having seen and discussed the result of this,
they may then replay the scenario with instructions
giving them more liberty to innovate.

Immersive Role-Playing as a Means of Prediction

Suppose an immersive role-play is based on accurate
analysis of a real-world situation. Then we study the
forces at work in the real world by studying role-player
behavior. This can yield predictions of a sort, as when
we find that a character seems to have no alternative
other than to comply, or vice versa, when we find that
there are no pressures on it to do so.
We cannot reliably get such predictions simply by
running a role-play and observing the results. Any


Confrontation Analysis

prediction of this sort should be subjected to criticism

and discussion aimed at overthrowing it. The roleplay itself should be used to draw attention to
questionable assumptions.
The point is that a drama-theoretic prediction, if
based on the presumption that there is only one way
to overcome a dilemma, such as by complying with
our position, itself generates pressures for its
overthrow because it motivates the character
concerned to find some other way. The only pure
predictions generated, even if our model and
briefings are accurate, are to the effect that
characters will face and attempt to eliminate certain
dilemmas. Further predictions are obtained by
asking: What kinds of dilemma elimination are likely
to succeed? Answering this question depends on
estimating the friction players will meet in different
directions, hence any useful predictions depend on
common-sense estimates of friction.

How Immersive Role-Playing Helps

How, then, could immersive methods help a
commander develop a confrontation strategy? We
suggest in the following ways:
Foreseeing likely reactionsRole-players
behavior can be used to foresee possible realworld reactions. Firm predictions of this kind
should be subjected to criticism of assumptions
before being accepted.
Encouraging criticism of key assumptions
Dilemmas can inspire role-players not only to

Chapter 10


change their briefings but to criticize the

assumptions made in them.
Asking key questionsAssumptions made
about beliefs and preferences in the briefings
given to role-players are key to the development
of a confrontation. Questioning them raises key
questions for a commander to answer.
Investigating assumptionsAssumptions
exposed to criticism can be investigated both
theoretically (i.e., asking What dilemmas are
strengthened or weakened if this assumption is
varied?) and in real life (i.e., intelligence can be
directed to answering specific questions about
other parties beliefs or preferences).
Understanding the implications of a
confrontation analysisA complex analysis
can be hard to digest. Role-playing gives
intuitive understanding.
Understanding intuitively the gradient of a
moment of truthThe term gradient is used
by analogy with a physical system, where it is an
object showing the directions a system is tending
to move. The drama-theoretic gradient of a
moment of truth is an object that sums up its
tendency to change as characters attempt to
eliminate dilemmas. Intuitive understanding of
the gradient of a confrontation is what a
commander needs to orchestrate the entire
compass of beliefs, attitudes, and emotions to tilt
the gradient toward his position.


Confrontation Analysis

Rehearsing real-world interactionsWe

perform better in a situation after having
developed skills and intuition through rehearsal.
Enhancing unity of effort across different
commands and levelsJoint role-playing by
those responsible for operations at different
commands and levels can generate common
understanding of the confrontation, enhancing
unity of effort.
Appreciating the viewpoints of other
participantsAn immersive briefing does not
take a single viewpoint, nor even a neutral one.
It takes each viewpoint and its viewpoint on
each others viewpoint andso forth; thus it
provides a conceptual platform from which to
survey all the viewpoints.

Mediation Support
Unilateral Decision SupportGeneral Considerations
Most of our discussion of how to win an OOTW has
assumed that the requirement is to support the
decisions of one character, ourselves, in confrontation
with others. In other words, we have been assuming
that our confrontation analysis is in unilateral mode.
What does this imply? First, that all information we
use is collected and kept by ourselves, except when
we selectively decide to release it to others. For
example, we may construct, for role-playing purposes,
a version of each characters viewpoint and its
viewpoint on others viewpoints, and so forth; however,
all these viewpoints are based on our own information,

Chapter 10


we do not go directly to other characters and ask for

their viewpoints.
Second, unilateralism implies that our strategy is
similarly confidential. We tell others only those parts
of it we wish them to know. They know the messages
we send them, or should know them. Messages sent
by physical action, such as bombing or delivery of aid,
should be accompanied by clear explanations lest they
be misinterpreted and send a message that was not
intended. Such clear explanations do not amount to a
full disclosure of our confrontation strategy.
Unilateral mode is the default mode for players that
neither belong to a common organization nor have an
acknowledged common purpose. Each must then
have a strategy for pursuing its ends and must send
messages to others as part of its strategy. Its ends
may be selfish or benevolent, atomistic or communal,
but they are its own ends.
However, we saw in chapter 3 that the very process of
conflict resolution between separate parties tends to
generate, through the emotions and arguments they
develop in trying to influence each other, a common
purpose and a common organization to support that
purpose. We sometimes express this by the phrase,
every drama is a character. This is how nations develop
and private revenge is replaced by public justice.
The development of law is only one example (although
an extreme and important one) of a general process
by which organizations develop formal and informal
procedures for conflict resolution. The essential
characteristic these procedures share is that they are
public and open, dependent on every step taken in


Confrontation Analysis

them being common knowledge among the players,

whereas the formation and implementation of a
confrontation strategy is essentially private. Even
telling other characters my confrontation strategy does
not make it common knowledge because saying that
it is my strategy may be merely part of my strategy. I
may be lying.
By contrast, many structures and procedures of an
organization must be public or they defeat their
purpose. It makes no sense to have a secret capital
city or a secret constitution. Many of these public
organizational procedures have an important conflictresolution function. This is obviously true of arbitration
and appeal procedures. Command procedures resolve
conflicts by specifying which officer has to make a
decision when there are conflicting views.
Informal conflict-resolution procedures are more tricky.
Suppose an office makes decisions by consensus,
but a new recruit does not know that and expects the
manager to decide on his own. This example illustrates
both the need for procedures to be common
knowledge if they are to be effective and the fact that
informal procedures are free to change and develop
with the players that operate them.
It is important to realize that each character, in
choosing how to go through the public steps of an
open conflict resolution process, is still pursuing its
own, private, intuitive confrontation strategy. (Here by
intuitive, we mean naturally developed, not derived
from knowledge of drama theory or confrontationanalysis methods.) Realizing this, we see that the
unilateral mode of implementing a private confrontation
strategy is not replaced by open institutions for conflict

Chapter 10


resolution. It continues to be pursued within them. The

difference they make is that each private strategists
choice of cards to think up and play is constrained by
the rules of the institution it is playing within, because
to break these rules would sometimes incur
punishment, at other times would be self-punishing in
diminishing the impact of the common-interest
arguments the strategist is using to make others accept
its position.
Thus, unilateral mode continues universally, even within
conflict-resolution procedures; however, it is often seen
as a Machiavellian pursuit of selfish interests to the
detriment of the common purpose. Why?
We all have a common interest in strengthening the
conflict-resolving institutions in our society and
influencing them to develop in ways that advance
common ends. Clear, intelligent development of
private confrontation strategies is thought to endanger
these institutions by showing how to take advantage
of them.
A generally excellent book on how to negotiate, Fisher
and Urys Getting to Yes (Fisher and Ury, 1983), heads
its first chapter, Dont Bargain Over Positions. This
would seem to contradict the whole of drama theory.
On examination, it does not. To begin with, the authors
recommend that each negotiator should solve its
deterrence dilemma, if any (i.e., have a fallback
position that puts pressure on the other). If the situation
is not to escalate, it follows that they must solve their
inducement dilemmas by converging to a common
position. The authors implicitly assume this. Given it,
the essence of their advice is, Dont talk about
positions, use rational arguments in the common


Confrontation Analysis

interest. Within the given structure, that is excellent

drama-theoretic advice. Not talking about positions
does not mean they do not exist. Rational arguments
in the common interest depend on the fact that for
each negotiator, even the others position is preferable
to the threatened future, avoiding which is a priority
for both. The arguments depend on finding the
common interests responsible; hence, the arguments
implicitly but clearly refer to positions.
In general, the animus against private strategies leads
to much hypocrisy. It is as if unilateral mode is seen
as part of our sinful nature. It cannot be avoided or
denied, but must be discouraged by being starved of
intelligent development.
We suggest that this attitude is partly correct.
Unintelligent, cynical expositions of how to form and
implement private strategies do undermine societal
and organizational conflict-resolving institutions;
however, drama theory shows how these very
institutions develop out of the pursuit of private
strategies, as private strategists find a need to argue
for the common interest. Drama-theoretic
confrontation strategies should strengthen, not
undermine, the institutions.
Meanwhile, the animus against private strategies
exists. It means that unilateral mode often will be seen
as appropriate only when parties are not part of a
common organization or are not pursuing an
acknowledged common purpose. Even then, it will
seem appropriate only if we approve of the purposes
our side is pursuing. Fortunately, these conditions are
met if we limit ourselves, as we have largely done, to
confrontations between our defense forces and

Chapter 10


potential rebels against the New World Order. This is

the post-Cold War world described in chapter 1.

Mediation Mode
We now ask, Can we use drama-theoretic insights
into the private strategies pursued within open conflictresolving procedures to improve the conflict-resolving
procedures themselves?
The answer is that we can. Confrontation mediation
(i.e., doing confrontation analysis in mediation mode)
involves doing a confrontation analysis that is seen
by and is common knowledge for all characters. The
aim is not to help one side to win, but to help the
process of conflict resolution itself. We want to move
it in a positive direction, toward a cooperative, happy
ending, rather than negatively, toward a tragic one.
The following question arises: Knowing that parties
are pursuing private strategies, how can we trust
the information they submit to an open, commonknowledge confrontation analysis? We will answer
this and other questions after outlining the
procedure, giving a simple example to make the
discussion concrete.

How Confrontation Mediation Is Done

Step OneAnalysts go to each side and interview
those involved. Views are solicited and carefully noted,
without criticism. This is important, not only to build a
good model but so that participants realize their views
have been incorporated (i.e., the model is not biased
against them).


Confrontation Analysis

Step TwoThe confrontation is analyzed using data

from Step 1. A model is built. There are now two cases,
as follows:
The model shows full and harmonious
agreement. It is presented to the parties. If they
confirm the agreement, the process ends;
otherwise, it goes to Step 3.
The model contains dilemmas. It is presented to
the parties, perhaps through immersive roleplaying with fictional names substituted for actual
ones. In discussions, dilemmas are high-lighted
and dilemma elimination encouraged through
positive discussion and arguments in the
common interest.
Step ThreeParties confer among themselves to see
if they can confirm with their internal subcharacters
any changes in attitude or beliefs generated at Step
2. The process then goes back to Step 1.
For example, suppose a memorandum of
understanding must be negotiated to provide for
cooperation between two forces, ourselves and
another. Our team at the negotiations (call them
representatives) want to make commitments that those
responsible for fulfilling the commitments (call them
suppliers) see as extreme, but representatives see
as necessary. Each side, representatives and
suppliers, has a reasonable case, but sees the other
as unreasonable. Each side is interviewed.
Step OneIn interviews, representatives give
excellent reasons for the need for the commitment,
suppliers for the uncertainty of being able to fulfill it.
Neither side will budge.

Chapter 10


Step TwoThe simple model in table 25 (first three

columns) is built and presented to the parties. After roleplaying they converge to the compromise in column C.
Representatives agree not to make the commitment
formally because suppliers are unsure they can fulfill it,
while suppliers agree to make every effort to fulfill it,
although they cannot guarantee success.
Step ThreeEach side confirms its changed position
Step One, round twoInterviews reveal that each side
means to keep the agreement, but distrusts the other.
Step Two, round twoThe new model is presented.
Each sides trust dilemma is discussed and eliminated.
Step Three, round threeNewly acquired trust in
the other side is confirmed internally, reported on
return to Step One, round three, and reconfirmed
by each party to the other at Step Two, round three.
End of process.

Harnessing Private Strategies to the

Mediation Process
Let us now answer the question, How can modelbuilders in this process trust the information they are
given in interviews, knowing that each party will be
trying to influence the process in its favor?
By choosing the right party to trust for each piece of
information, a mediation model can be made
deception-proof in the following sense: each partys
attempts to influence the process in its favor will help
the process rather than hinder it.


Confrontation Analysis

To show how this works, we will use as an example

the process by which mediation resolved the simple
confrontation between representatives and suppliers
in table 25. In the first model, constructed in the first
Step 2, each side has an inducement dilemma
(preferring the others position to t) and a threat
dilemma (preferring not to carry out its threat, if the
other does).
Now observe that it is in neither sides interest to admit
that it faces these dilemmas; therefore, if either of them
refuses to admit to a dilemma, we can point out that
what matters to it is not, actually, whether it has the
dilemma, but whether the other believes that it has.
Seeing this, it can be encouraged to produce rational
common-interest arguments and give evidence to
convince the other it does not have the dilemma. Its
use of common-interest arguments in particular can
be encouraged by pointing out their greater
effectiveness in influencing the others beliefs, which
is what it needs to do.
If, on the other hand, either party cannot believe that
the other is unconvinced, the purportedly unconvinced
party is shown that what matters to it, again, is not
that it is unconvinced, but that the other should believe
that it is; therefore, it can be encouraged to use rational
common-interest arguments and produce evidence as
to why it ought to remain unconvinced.
This general principle, of laying the burden of
producing conviction on the one who needs to do so,
and suggesting how it should be produced, will, if
followed consistently, bring about convergence to
common beliefs through the use of reason and
evidence, passionately argued.

Chapter 10


Table 25. Confrontation between representatives and suppliers.

Moreover, in this process of convergence, characters

will have made appeals to and explorations of their
common interests. Their motivation will have been
to convince each other, but the effect will be to
construct a view of themselves as having interests in
common that will prompt them to think of win-win
solutions and compromises.
Continuing with our example, suppose that
representatives and suppliers have found a
compromise solution and in Step 2, round two, found
that they cannot trust each other. Each is told that its
problem is not whether it is trustworthy, but that the
other should believe that it is. As before, it needs to
produce rational common-interest arguments and
evidence to convince the other, and may in the process
convince itself, if it had in fact intended to defect, that
defection is against its interests.
In this way, the process again brings about
convergence to common beliefs with accompanying
reframings of the situation that actually may change


Confrontation Analysis

perceived facts, as well as beliefs. The general

principle to be observed is that the unbelievability, for
good structural reasons, of a characters evidence can
be used in the mediation process to persuade that
character to provide better arguments and evidence,
based on the parties common interests, simply
because it is always other parties, not itself or the
mediator, that it needs to convince.

When Should It Be Used?

Theoretically confrontation mediation is usable
whenever parties are willing to participate; however, it
must be done by external experts, who require funding.
The party that funds these experts naturally lays itself
open to the suspicion that it is pursuing a private
strategy. Such suspicions are certainly aroused by
attempted mediation in many OOTW confrontations.
To overcome this problem in a fundamental manner,
confrontation mediation, like established conflictresolution procedures such as legal systems, needs
to be funded by an organization to which all parties
belong and owe loyalty.
This is because the above suspicion is justified. The
funding organization must be pursuing a private
strategy; we all are. The question is, with what aim?
The only answer that will satisfy the parties is the
general, long-term, common interest of the parties
themselves. To ensure this aim requires an institution
such as the law, that is owned by an organization (here,
the nation) to which all parties belong, yet is kept
separate from its policy institutions (because policies
may be what the confrontation is about).

Chapter 10


Arrangements like these can be made within an

organization such as a firm or government agency. A
mediation unit can be set up with guarantees of
independence. It might even be done within a coalition
of defense forces assembled for a specific OOTW.
Our argument is that unless it is done, confrontation
mediation will be used at most sporadically.

Informal UseBeing Your Own Mediator

When formal mediation is impracticable, the ideas of
mediation may be of use to those implementing their
own, private confrontation strategies.
There is a sense in which I need to be my own
mediator. If I wish to avoid tragic outcomes, as by
definition I should, then I want the process as a whole
to go in a positive direction, in addition to wanting it to
go in my own particular direction. I can help this to
happen by observing myself and other parties in
mediation mode and giving mediation-type advice to
myself and them as needed. Familiarity with the
principles of mediation support is useful to all involved
in confrontations.
Another consideration is that confrontations vary along
a continuum of expected cooperativeness according
to the degree to which cooperation is expected or
demanded of them. This is not a matter of how
characters feel. Often parties in the most hopefully
cooperative relationships have the strongest negative
feelings, just because they cannot cooperate. Vice
versa, parties trying to kill each other may feel very
little toward each other, or may even feel friendly.
Emotions, as we have seen, have particular functions


Confrontation Analysis

at a moment of truth within a confrontation; they do

not make confrontations cooperative or not.
That is brought about by organizational relationships.
Parties are expected to cooperate when they are in
functional relationships within the same or related
organizations. Hostile armies confront each other noncooperatively in our sense, whereas armies in a
coalition, who may frustrate and infuriate each other
as much or more, do so even though their
confrontations are hopefully cooperative.
It is this variable factor of expected cooperativeness
that seems to determine how strongly private
confrontation strategies are deplored, regarded as
disruptive, and denied funding. Organizations deplore
Machiavellian behavior within their ranks. We might
hope to use mediation support instead of unilateral
confrontation support to deal with such problems, except
that entrenched positions and angry attitudes often
found in hopefully cooperative confrontations may make
it hard to get participation. In such cases, being my
own mediator may be the best solution available.

Summary of Chapter 10
This chapter deals with two topics not covered
elsewhere in the book. Immersive briefing is a dramatheoretic tool for briefing role-players to enable them
to take the parts of characters in a confrontation that
has been analyzed. Each role-player is told about its
characters background, values, projects, relationships
with other characters, and confrontations with others.
In relation to each of its confrontations, it is told who
the characters involved are, what their positions and

Chapter 10


fallback positions are, and which tactics they are using.

Each character in a drama gets its own briefing,
different from the others. A character may be roleplayed by a team or an individual.
Briefings are composed using a software tool and
distributed as software packages, through which
information is accessed by clicking and pointing. After
being briefed, role-players interact with each other to
try to resolve their problem. Interactions can be via
various media, from e-mail to personal meetings.
Immersive role-playing can help a commander to
develop a confrontation strategy by enabling him at
step 2 to test out the likely effects of his strategy on
other players. The results of such testing need to be
thoroughly criticized and discussed.
Mediation support is contrasted with unilateral decision
support as an alternative mode of using confrontation
analysis. Instead of being used to develop a private
strategy for one player, confrontation analysis in this
mode is used to model a problem for the equal benefit
of all players. Information is obtained from all and
distributed to all. The modeling process is used to help
the conflict resolution process move toward a happy
ending rather than a tragic one.
It might seem problematic to rely on the information
supplied by each participant in confrontational
mediation because each will want to give information
that will bias the procedure toward its own ends;
however, this is not a problem because this private
motivation of each participant can be used to drive
the process forward. This is done by pointing out to a
participant whose information is doubted that it is not


Confrontation Analysis

the truth of its information so much as its acceptance

by others that matters to the participant. In this way,
the participant is encouraged to use rational arguments
in the common interest to convince others.
Mediation mode may be considered most appropriate
to problems where parties belong to a common
organization or have acknowledged common ends,
whereas unilateral mode may seem right only when
these conditions are not met; however, unilateral mode
also encourages happy endings, if used intelligently.

Alberts, D.S., & R. Hayes. Command Arrangements
for Peace Operations. Washington, DC:
National Defense University Press. 1995.
Bennett, P.G. & Howard, N. Rationality, Emotion and
Preference Change: Drama-Theoretic Models
of Choice. European Journal of Operational
Research. 92: 1996. 603614.
Bennett, P.G. Toward a Theory of Hypergames.
OMEGA 5(6): 1977. 749751.
Bryant, J. All the Worlds a Stage: Using Drama
Theory to Resolve Confrontations. OR Insight.
10(4): 1997. 1421.
Bush, G. Speech by U.S. President in September 1990
following Iraqs invasion of Kuwait. 1990.
von Clausewitz, C. On War. Edited with introduction
by A. Rapoport. London: Penguin Books. 1968;
1st edition 1832.
Davies, N. Europe: A History. London: Pimlico. 1996.
DERA. Drama-Theoretic Tools for Operational
Analysis. Report under contract number CDA/
P/564. 1997.
Fisher, R. & W. Ury. Getting to Yes . London:
Hutchinson. 1983.



Confrontation Analysis

Howard, N. Paradoxes of Rationality:Theory of

Metagames and Political Behavior. Cambridge,
MA: M.I.T. Press. 1971.
Howard, N. Drama Theory and Its Relation to Game
Theory. Part 1:Dramatic Resolution vs. Rational
Solution. Group Decision and Negotiation. 3:
Howard, N. Drama Theory and Its Relation to Game
Theory. Part 2: Formal Model of the Resolution
Process. Group Decision and Negotiation. 3:
1994a. 207235.
Howard, N. Negotiation as Drama: How Games
Become Dramatic. International Negotiation.
1: 1996. 125152.
Howard, N. What Is Drama Theory? In Modelling
International Conflict: Proceedings of a
Conference by the Institute of Mathematics and
Its Application. Southend-on-Sea: IMA. 1997.
Howard, N. Drama Theory: The Fundamental
Theorems in Modelling International Conflict:
Proceedings of a Conference by the Institute
of Mathematics and its Application. Southendon-Sea: IMA. 1997a.
Howard, N. N-Person Soft Games. Journal of
Operational Research Society. 49: 1998. 144150.
Howard, N., P. Bennett, J. Bryant, and M. Bradley.
Manifesto for a Theory of Drama and Irrational
Choice. Journal of Operational Research
Society. 44(1): 1992. 99103.
Joint Chiefs of Staff. Joint Vision 2010. 1997.



Kahn, H. On Escalation: Metaphors and Scenarios. New

York, Washington: Frederick A. Praeger. 1965.
Kuhn, T. S. The Structure of Scientific Revolutions.
Chicago: University of Chicago Press. 1962.
Maxwell, D. Adaptive Command and Control: Future
Mission Statements in Peace Operations.
Proceedings of Third International Symposium
on Command and Control Research and
Technology, National Defense University,
Washington, DC. 1997.

Cooperationor Conflict (Newsletter). Birmingham,

UK: Nigel Howard Systems. 19921997.
Osborne, M., and A. Rubinstein. A Course in Game
Theory. Cambridge: MIT Press. 1994.
Owen, D. Balkan Odyssey. London: Indigo (Cassell
Group). 1996.
Popper, K. The Logic of Scientific Discovery. London:
Hutchinson. 1959, original German edition 1934.
Rapoport, A. Fights. Games and Debates. New York:
Harper. 1960.
Rapoport, A. Strategy and Conscience. New York:
Harper. 1964.
Rapaport, A. Preface to Clausewitz On War. London:
Penguin Books. 1968.
Rogers, P. War in the 21st Century: Likely Trends and
Their Possible Control. In Modelling
International Conflict: Proceedings of a
Conference by Institute of Mathematics and Its
Applications. Southend-on-Sea: Institute of
Mathematics and its Applications. 1997.


Confrontation Analysis

Schelling, T.C. The Strategy of Conflict. Cambridge,

MA: Harvard University Press. 1960.
Silber, L., and A. Little. The Death of Yugoslavia.
London: Penguin Books (BBC Books). 1996.
Tait, A. Drama Without Tears. Tutorial Papers, Young
OR 10. Birmingham, UK: Operational Research
Society. 1998.
von Neumann, J., and O. Morgenstern. Theory of Games
and Economic Behavior. Princeton, NJ: University
of Princeton Press. 1959; 1st edition 1944.
Wentz, L.K. C3I for Peace Operations: Lessons from
Bosnia. Proceedings of Third International
Symposium on Command and Control
Research and Technology. Washington DC:
National Defense University Press. 1997.

About the Author

igel Howard is Managing Director and Chairman

of ISCO Ltd and Visiting Professor at Sheffield
Business School, Sheffield Hallam University, UK. His
early work in game theory focused on the practical
dilemmas created when one tries to take a purely
rational approach to human interactions. His work
has remained practical, while remaining focused on
this same problem. His methods were applied, in the
late sixties, to the first SALT agreement, nuclear
proliferation, and Vietnam; they have since been used
on many business and governmental policy problems.
In recent years he has been the chief developer of a
new approachdrama theory. He has taught at the
LSE and the universities of Pennsylvania, Waterloo,
Ottawa, and Aston. From 1987 to 1997 he published
a bimonthly research letter, Cooperation or Conflict,
now an Internet site (\
drama.htm). From 1997 onwards he has applied a
technique derived from drama theoryConfrontation
Analysisto defense problems, in particular to Peace
Operations. Currently he is engaged in a project for a
Command and Control System for Confrontations. In
the non-defense sector, he is involved in work on both
mergers and acquisitions, and also in designing a
virtual hospital for medical personnel training.